MARINE BIOLOGICAL LABORATORY.

Received June 20t 1938

Accession No. 49112

Given by MoGraw-Kill Book Co.,I^io«

Place, New York City

***No book op pamphlet is to be removed from the Lab-
oratory without the permission of the Trustees.

M. CRAW -II I I.I. PUBLICATIONS 1 X Til I

ZOOLOGICAL SCIENC] S

A I RANKLIN SHULL, Cowsi wwg Editob

PHYSIOLOGICAL GENETICS

Selected Titles From

McGRAW-HILL PUBLICATIONS IN THE
ZOOLOGICAL SCIENCES

A. Franklin Shull, Consulting Editor

Chapman Animal Ecology
Goldschmidt ■ Physiological Genetics
Graham • Forest Entomology
Haupt ■ Fundamentals of Biology
Metcalf and Flint ■ Insect Life
Mitchell ■ General Physiology
Pearse ■ Animal Ecology

Reed and Young ■ Laboratory Studies in Zoology
Riley and Johannsen ■ Medical Entomology
Rogers ■ Textbook of Comparative Physiology
Senning ■ Laboratory Studies in Comparative Anatomy
Shull ■ Evolution
Heredity
Shull, LaRue, and Ruthven ■ Principles of Animal Biology
Snodgrass ■ Principles of Insect Morphology
Van Cleave ■ Invertebrate Zoology
Welch ■ Limnology
Wieman • General Zoology

An Introduction to Vertebrate Embryology
Wolcott ■ Animal Biology

There are also the related series of McGraw-Hill Publications in
the Botanical Sciences, of which Edmund W. Sinnott is Consulting
Editor, and in the Agricultural Sciences, of which Leon J. Cole is
Consulting Editor.

PHYSIOLOGICAL
GENETICS

BY

RICHARD GOLDSCHMIDT, Ph.D., M.D.. D.Sc

Professor of Zoology, University of California

I I KM KlMTION

Mi GRAW-HILL BOOK COMPANY, [nc.

N I . w fORK AND LONDON

i g

Copyright, 1938, by the
McGraw-Hill Book Company, Inc.

PRINTED IN' THE UNITED STATES OF AMEKH'V

All rights reserved. This book, or

parts thereof , may not be reproduced

in any form without permission of

the publishers.

THE MAPLE PRESS COMPANY, YORK, PA.

PR K PACK

This is the third time thai I have presented the sub j eel matter
of this book. Each time it has been in differenl form. The first
little book, "Die quantitative Grundlage von Vererbung und
Artbildung," written in L917 and published in 1920, consisted of
a number of essays in which I developed my views on the prob-
lems of gene and development as derived from my own work on
Lymantria dispar. There was not much material from other
sources to be quoted. Ten years later I returned to the same
subject. Meanwhile my own work had amplified, experimental
embryology had entered a new phase, and a number of other
geneticists had become interested in the problems of genie action.
Notable among these were S. Wright, whose interest in the
subject dates back as far as 1916; Zeleny and his school, who
furnished most exact quantitative data; J. S. Huxley, who had
contributed to the theoretical analysis and had started experi-
mental work; Sturtevant, who had just cleared up the case of
Bar eye in Drosophila; and Plunkett, who had applied his
physicochemical knowledge to certain problems. This second
book, "Physiologische Theorie der Vererbung," was meant to
elaborate in all details my own views regarding the action of the
gene in development , but it also reported whatever relevant work
of others was available. Again ten years have passed, and these
have witnessed an ever-growing interest in the field and a corre-
spondingly increasing amount of work, which, with the introduc-
tion of new methods, is expanding more and more. As it is
emphasized over and over again by writers of texts and by general
speakers that we know next to nothing of the action of the
hereditary material in controlling development, it seems advis-
able to present the entire material available. Not only has this
material been assembled and reviewed, but an attempt has been
made to organize it into the skeleton of a future science of

m PREFACE

physiological genetics. Needless to say, the reader is Bupposed
to have sufficienl knowledge of genetics and embryology.

The author takes pleasure in expressing his gratitude to
Professor E. B. Babcock, who was so kind as to go carefully over

the whole manuscript; also to those publishers who permitted
me to copy figures.

Richard Goldschmidt.
(M\ BBS] rv of California,
Berkeley, Calif.,
January, 1938.

CONTENTS

Paoi

Preface v

I. [introduction !

II. The Mutated Gene \m> the Potentialities of Development 3

1. The phenocopies 4

A. Description 4

B. Related phenomena 10

C. The sensitive period 13

D. Combination effect of gene and phenocopy 22

2. The phenocopies and the development of mutants 23

A. Facts of mutant development 23

B. Mutant character and developmental stage 39

C. Lethality 42

D. Preliminary conclusions 44

E. Some pertinent facts from experimental embryology . ... 46

3. The mutant gene and the rate concept 51

A. Rate genes 52

B. General experimental evidence 56

1. A typical case 56

2. The threshold concept 65

3. Penetrance 69

4. Further evidence 72

4. Some special types of mutant genes and genie actions 78

A. Pleiotropy 78

B. Multiple alleles 81

C. Genes controlling known chemical processes 90

5. The gene in heterozygous state 99

A. The influence of the environment upon dominance .... 102

B. Dominigenes 107

C. Change of dominance 112

I). Dominance and chemistry of products of gene-action . . . 113

/•.'. Dominance in multiple allelomorphic series 114

/-'. The phylogenetic theory of dominance 121

6. The gene in different doses 123

A. Differences of dosage produced by the IX— 2X- condi-

tions 125

1. Sex-controlled phenotype 125

2. Dosage through Bex-chromosomes 130

B. Differences of dosage by deficiencies 133

C. Differences of dosage by addition of genes 135

vii

4911'

viii CONTENTS

Page

I). Differences of dosage no! confined to individual genes I 17

/'- The heterozygote as dosage difference 154

7. The interaction of the genea 156

8. The type of reaction controlled by the tnutanl gene .... 171

A. Reactions within the cells 171

B. React ions of an induct ive type 181

1. Terminology lsi

2. Genie products of the hormone type 182

3. Diffusion of growth— or similar substances 190

4. The determination stream 193

9. The problem of pattern 200

A. Pattern in development 200

B. Genes controlling developmental processes 205

C. Genes and pattern 209

1. Patterns of growth and form 210

2. Symmetry 221

3. Special patterns superimposed upon the general pattern 228

4. Composite patterns 241

a. General embryological features 241

b. Pattern localization 245

c. Additional facts 255

III. The Cytoplasm and the Activation of the Gene 263

1. Interaction of nucleus and cytoplasm 263

2. The activation of the genes 265

3. Cytoplasmic heredity 267

A. Merogony 267

B. Plastids 269

C. Cytoplasmic influence upon gene-controlled characters. . . 270

D. The theory of the plasmon 275

E. Cytoplasmic inheritance and Dauer-modification 276

F. Cytoplasm and sex 278

G. Genes controlling cytoplasmic properties 280

4. Conclusions 280

IV. The Nature of the Gene 282

1. The gene as an active molecule or group of molecules 283

A. Mutation as change in quantity 285

B. Mutations as changes of side chains or residues 288

C. Mutations as changes of a stereoisomer^ type 292

D. Mutations as polymerizations 292

2. The theory of the gene as based upon physical deliberations 293

3. The gene as an aggregate of different parts 295

A. The genomere theory of Anderson and Eyster 295

B. The theory of the Serebrowsky school 295

C. The hypothesis of Thompson 297

4. The gene as loss of chromosome material 298

5. The gene as part of a higher unit 299

CONTENTS ix

Paoi

1 The gene ae a Bide chain of a molecule 300

H. ETriesen's theory of chain mutation 302

6. Outlook upon the theory of the gene of tomorrow 303

A . The position effecl :!(| I

B. Interpretation in terms of genes 308

C. Is the theory of the gene still valid? 309

Bibliography 317

Aithok Index 363

Subject Index 369

PHYSIOLOGICAL GENETICS

I. INTRODUCTION

Genetics is the study of heredity. As the causation of all
hereditary traits must be found in the fertilized egg (or its
physiological equivalent in protista and lower plants), it is the
aim of genetics to link the agents within the germ cells with the

hereditary traits observed in the fully developed individual.
There1 are two main aspects of this problem: first, the unknown
agents in the germ cells have to be identified; their mode of
transmission in all its phase- has to be ascertained; and the facts
found have to be brought in line with the results of genetic
experimentation. As is generally known, the study of these
problems has accumulated the bulk of facts usually considered
as genetics proper, culminating in the generally accepted theory
of the gene as the hereditary unit and the chromosomes as its
seat. This side of the problem of heredity may be called the
problem of the mechanism of heredity, and the work that has
solved it is the main content of what is usually called genetics.
There is a second aspect of the problem of heredity: to under-
stand how the gene, whatever it is, acts in controlling typical
development to the adult form showing all the hereditary traits.
One might call this problem simply the problem of development.
But as development is to be linked specifically with the function
and action of the genes, this special chapter may be termed the
physiology of heredity; and the science devoted to its solu-
tion, physiological genetics. ( >ne might also speak of static genetics,
studying the status of the germ plasm; and of dynamic genetics,
inquiring into the acting forces that lead to the visible effect.

The problems <>t' static genetics may be and have been attacked
directly, and the laws and rules that have been derived from
experiments in this field may actually be made visible in con-
comitant cytological studies. Physiological genetic-, however,
does not offer any method of direct attack and therefore calls

1

2 PHYSIOLOGICAL GENETICS

(•(instantly for interpretative hypotheses which are sometimes
difficull to prove. This constitutes simultaneously the danger
and the charm of the work in this line. The effect has been thai
comparatively few geneticists were attracted to this difficult
field and that those who were had to labor in the dark. Thus,
the part of genetics that to me is the most interesting was
practically banned from advanced treatises and textbooks of
genetics, and the opinion has developed and has even been voiced
that it is not worth while to mention a field in which nothing is
known with certainty. But the history of science teaches us
that it is just such borderline fields of today that will bear the
fruits of tomorrow. Fortunately the negative attitude has begun
to change considerably, and more and more workers of the younger
generation are attracted to the problems of physiological genetics.
It might therefore be useful to put together the known facts and
to link them into a picture, knowing that this picture will be
subject to many changes and improvements even in the near
future.

II. THE MUTATED GENE AND THE POTENTIALITIES
OF DEVELOPMENT

In the following discussions, we take the existence of the gene
for granted without further inquiry into its nature, as the facts
that we shall have to report and the conclusions to be drawn arc
largely independent of the conception- regarding the nature
of the gene. One of the basic facts of genetics is, then, thai the
action of the gene in controlling the typical development of
hereditary traits cannot be studied directly but may only be
extrapolated from the knowledge of the action of a mutant gene:
the existence of the normal or +- gene is only inferred from the
existence of a mutant allelomorph showing Mendelian behavior.
The action of a mutant gene is a different one from the assumed
action of the -f— gene; i.e., the mutant gene must control or
produce a deviation in the series of developmental processes
leading to the visible character. Development, as we know, is,
leaving aside the difficult problem of regulation, a process of
extraordinary precision. Under constant conditions, one step
follows the other with almost invariable precision in regard to
space and time. Each step is dependent upon the normal appear-
ance of the preceding one, and the normal result depends —
barring regulations upon the orderly sequence of events in
quality, quantity, time, and space. One of the consequences of
this situation is that the possibilities of changing the details of
developmental processes without injuring the proper cooperation
of those processes are rather limited. Most of the changes of
individual processes that might be produced will throw out of
gear the combined system, and the result will be destructive.
But certain processes may be changed without deleterious
consequences; and if tin- i> done by a genetic change, we call it
a mutation. Such considerations, obvious as they seem to be,
make US expect that the action of mutated scenes upon develop-
ment cannot he of a different type from any other changes of

development induced by experimental agencies: in both cases,
something changes the detailed course of some developmental

3

4 PHYSIOLOGICAL GENETICS

process or processes; and in both cases, the quantity and quality
of the possible change are limited by the degree of freedom left
to the individual developmental processes without destroying
their integration into :i more or less normal whole.

These considerations show that it is <>i primary importance
for an understanding of the action of the gene to compare tho
effects of the mutated gene upon development with effects
produced by external agencies upon the development of the Wild
type.

1. THE PHENOCOPIES

A. Description

In his classic researches upon the wing pattern of butterflies,
Standfuss (1896) showed that it is possible to change this pattern
by the action of extreme1 temperature in the pupal stage so that
the experimentally produced pattern cannot be distinguished
from the pattern of known races of the same species; e.g., heat
treatment of the Swiss Papilio podalirius could produce a form
looking like the variety P. zanclaeus found in Sicily. Central
European Papilio machaon pupae treated with heat gave rise
to butterflies resembling the Syrian race P. sphyrus or the
Turkestan race P. centralis. Central European Vanessa urticae
pupae treated with low temperature produced types like the V.
polaris from Lapland ; and the same treated with heat gave forms
indistinguishable from the race V. ichnusa of Sardinia (Fig. 1).
In this case, the geographic races that were copied in the experi-
ment were known to be constant and therefore presumably
genetically different from the main form, though this genetic
difference \vas not established.1 The facts were used mostly
for phylogenetic and Lamarckian speculations, and only in
1917d, 19206, and 1927c did Goldschmidt point out their impor-
tance for the understanding of the action of the gene. Recently
Goldschmidt (1935a) proposed the term phenocopy for such forms,
produced experimentally from the Wild-type form, the phenotype
of which copies or duplicates the appearance of a mutant (or
combination of mutants) of the same form. We shall use this
term henceforth.

1 In other experiments, heat treatment of female pupae produced the male
type of wing pattern. Here, of course, as we know now, the normal differ-
ence is controlled by genes.

THE MUTATED GENE

Fig. 1. — 1-8, eight different phenotypes produced by temperature action upon
the pupae of Vanessa iirtia, ; '.i, the usual type; 13, 14, the geographic race
ichnusa; 10, ichnusa-like type produced as phenocopy. ( fJh<iln(jrn/ili Fischer,
from Goldschmidt, Einf. i. il. Yvrerbywiss., 5th ed., 192i», Fig. 17 t.

6

PHYSIOLOGICAL GENETICS

I Ine of the elementary conceptions of classic genetics is, of
• oursc, that one and the same phenotype may be hereditary or

only a modification, and examples of this arc found in all lines
of genetic work. But very rarely has the question been raised

Fig. 1. — (.Continued.)

as to whether or not there is a physiological connection between
both phenomena, which might be attacked experimentally.
Some beginnings are found in the earlier work of Tower (1906),
who claimed that in Lcptinotarsa he could produce mutations or

THE Ml TATED GE \ I 7

their phenocopies, according to the way he treated hia animals
with abnormal conditions. This, however, has never been
verified. Heche's (1907) experiments changing the feathering
of the dove ScardafeUa inca hy action of moisture also belong
here. They resemble Standfuss' experiments in so far as the
type was shitted toward the type of a known but unanalyzed
geographic race

The first controlled experiments with genetically known char-
acters were made hy TimofeefT (1926), who reported that some

[nutations found in Drosophila funebris were paralleled by
identical mollifications. In one case, a change of temperature

COllld produce one of these phenocopies. Contorted wings.

A systematic study of this problem was inaugurated by
Goldschmidt, who reported in 1929a that it is possible to repro-
duce the phenotypic likeness of many mutants of Drosophila by
treating larvae of a definite stage with temperature shocks, and
this in a quite regular manner. The main results as published
later (1935a) are:

If Drosophila larvae of a definite age are treated with tempera-
ture shocks. of 35 to 37°C, a considerable number of the flies
hatching from such larvae are more or less modified. Among
these modifications we find many types of abnormalities and
simultaneously the phenocopies of numerous known mutations.
Among these are most frequently the changes of bristles, copying
practically all known types of bristle mutations, e.g., forked,
bobbed, stubble; further, the copies of all types of wing mutations
like miniature, arc, (airly, pointed, dumpy, scalloped; further, the
copies of many types of mutation of eye form like small eye, lozenge,
star; furthermore, rare but sufficiently frequent copies of differ-
ent mutants of body form and appendages like aristaless, daehs.
abnormal abdomen, and also such types as benign (tumor);
more rarely copies of changes in color and design like sooty,
t rident. All of these phenocopies appear not only in the complete
form represented by the well-known mutants but in a series
of different grades resembling all types of known multiple allelo-
morph- or such as are to be expected without having been found.
Generally speaking, one might say that all groups of known
mutants are represented and a considerable number of individual
members of each group. As only one method with not too many
variations was used, one might expect that a proper variation of

8 PHYSIOLOGICAL GENETICS

method would produce the phenocopies of all known mutants

(sec pages <) and 10). It ought to be added that in these experi-
ments the modified individual may exhibit simultaneously quite a
number of different traits; e.g., the wing might he simultaneously
short, spread, rolled, and even the two wings different.

The following details are of importance. Within a rather
considerable range of variation, typical effects are dependent
upon four variables:

1. Developmental stage at which the heat shock is applied.

2. Time of exposure to heat (6 to 24 hr.).

3. Intensity of heat (35 to 37°).

4. Genetic condition of the material.

It might be said that the number of phenocopies (up to 100 per
cent) and the degree of phenocopic change (up to the highest
member of a given series) are roughly proportional to the product
of the time and the temperature of the exposure, i.e., to the
intensity of the shock. The specific type of phenocopy produced
is dependent upon the intensity of the shock, the time of its
application, and the genetic line used for the experiment. It
would be surprising, of course, if so rough a method were to
produce a 100 per cent typical result for all different combinations.
But it is a fact that, at least for some of the phenocopies, a formula
can* be given which always produces them within a certain range
of variation. Table 1 contains some data as found in a defi-
nite series of experiments with the Oregon wild line.

It is further very typical that certain phenocopies tend to
appear together with a definite treatment. Thus it is possible to
produce quite regularly the combination of spread wings and
long (angora) bristles or of curved wings and short stubbly
bristles.

The influence of the race is a rather considerable one. Some
wild or mutant races respond more easily to the stimulus than
others. Some are more apt to produce phenocopies of the
bristles; others, with the same method, give more wing modifica-
tions. And some of the phenocopies were produced exclusively
or almost exclusively within a definite line. We shall return later
to this point.

A special interest attaches to the time element, the sensitive
or critical period at which the phenocopies may be produced.
This important point will be treated in a separate chapter.

THE MUTATED GENE 9

Further important facte have been added by other authors.
Jollos (1933) confirmed many of the foregoing findings in connec-
tion with work in a different line. Gottschewski (1034) found
that exposure to cold produced similar effects. This was to be
expected on the basis of the classic experiments on butterflies,
where Dorfmeister (1864, 1879), Merrifield (1889, 1894), Stand-
fuss (1896), and Fischer (1895) had shown that extremely low
temperatures produce the same modification- of the wing pattern

Table 1. — Pkouiction of Dbfinitb Phenocopibb undbb Typical

Conditions

Phenotypr

Scalloped
Curly ....

Ski

Spread . . .
Curved. . .
Dumpy . .
Lancet. . .
Miniature
Blistered .
Rolled
Trident. .
Eye size . .
Horns. . . .
Benign . . .

Age days

6-7

ti ti'.,

5H

5-7

5

7

V., 7

5-6

7

7

51,-r,

7

35

35-37

36-37

35

36

36

36-37

36-37

36

35-37

35-37

37

35

35

Exposure,

hours

12-24
18 24
12

18-24
18
12
18

12-18
18

18-24
C 24
18
24
18-24

( >|if i 1 ■ 1 1 1 1 1 1
per cent of
phenocopies

70
76
43
91
23
34
22
40
10
40
82
100
4
75

as high-temperature shocks. In Gottschewski's freezing experi-
ments also phenocopies ni the eye-color mutants appeared, which
had been absent in the heat experiments. Recently Friesen
(1936) reported the production of phenocopies in Drosophila
by the action of X rays. His results, as far as they go, agree in
all the points mentioned with those of the temperature experi-
ments. It seems, however, that some of the phenocopies
produced by temperature shocks do not occur in the X-ray
experiments, and vice vcrsi.

Here belongs also the report by Bobroff (1930) that X-ray
treatment of caterpillars results in males' showing some female
characteristics (see footnote, page 4). Geigy (1926) produced

10 PHYSIOLOGICAL GENETICS

by action of ultraviolet rays abnormal abdomen in Drosophila,
resembling the well-known imitation.

We have mentioned the fad that occasional phenocopies arc
found in many cases, apart from the systematic experiments just,
reported. One of these may be reviewed here because it belongs
to a case of special genetic and embryological interest. Another,
the experiments of Kuelm and Henke on the flour moth, will
be reported in the following chapter.

A well-known hereditary abnormality in fowl is rumplessness,
which has been studied genetically and anatomically by Du
Toit (1913), Dunn (1925), and Landauer (1928). It consists of
a nidi mentation of the end of the vertebral column (see page 26).
( Occasionally also in normal strains individuals are found who
exhibit this character in a nonhereditary form (about 1 in 1,000
individuals). Danforth (1932) succeeded in producing this type
as a phenocopy in 7}^ per cent of eggs treated with varying
abnormal temperatures during the first week of incubation. As
an explanation he accepts Stockard's views, which will be reported
on page 47.

There can be no doubt that similar facts will be found in plants.
Numerous examples are known of modifications resembling
hereditary types. No systematic attack upon the problem, using
known gene mutations, is known, though it might be easily done
for floral abnormalities, leaf forms and types, and the like.
A hint at such facts is found in Zimmermann (1934).

For curiosity's sake it might finally be mentioned that even
certain details of chromosome behavior fall into this group:
Genes are known that prevent synapsis of the chromosomes, and
the same phenomenon may be produced by action of abnormal
temperature (Oehlkers, 1936). There are genes causing abnormal
fertilization with consequent gynandromorphism (Goldschmidt-
Katsuki, 1927), and the same effect may be produced by tem-
perature shocks (Roesch, 1928).

B. Related Phenomena

Thus far we have recorded the cases in which the phenocopies
of different type were produced from the Wild type. There is
another group of experiments, also performed on Drosophila, in
which the phenotype of quantitatively different members of a

THE Ml TAT ED GE VE

II

-cries of multiple allelomorphs was reproduced I hrougfa treal menl
of a member of the series. There is first the work of Zeleny and
Krafka oe the Bar-eye Beries in Drosophila, Seyster (1019;
and Krafka (1920) had shown thai the size of the eye and the

number Of ommatidia in Bar-eye Drosophila vary with the
1500

15 20 25 30

Temperature in degrees C .

Fig. 2. — Data for homozygous females of various stocks of the Bar-series in
Drosophila plotted semilogarithmically. 1, Wild type full; l'. reverted full;
3, unselected Bar; 4 6, different low selected Bar; 7-9, different Ultrabar.
{^rom ITrsh, L930, ./. Exp. ZooL. 57, Fig. 1.)

temperature at which the larvae are developed. < >n the average
an increase of 1°C. produces ;i decrease of about 10 per cent in
ommatidia] number. In Ultrabar, the change is aboul 8 per cent ;
tui- Wild eye t he efYect is about 21 ■> per cent. In flies heterozygous
for Ultrabar ;iii<1 Wild, tin- percentage is equal to that ni Ultrabar.
(We shall return to this point in the chapter on dominance.)

More exact measurements were taken by Luce (1926) and
Driver (1920, 1931) for different temperature- and alleles. A- an

12 PHYSIOLOGICAL GENETICS

example illustrating our point we reproduce a representation of

Hersh's data for different lines of Wild, Bar, and Ultrabar at
different temperatures, plotting; temperature against the loga-
rithms of facet number (Fig. 2). The point of interest for the
descriptions in this chapter is the phenocopic effeet in both
directions between Bar and Ultrabar: at 15°, genetic Ultrabar
resembles genetic Bar at 25°; etc. Just as the Bar series of
multiple allelomorphs controls different numbers of ommatidia,so
temperature acting upon the larvae shifts the phenotypes within
the series of existing or possible allelomorphic phenotypes.

Still more interesting in regard to the analysis of gene action is
the work on temperature action upon the wing of the mutant ves-
tigial in Drosophila, inaugurated by Roberts (1918) and continued
by Harnly (1930-1935), Nadler (1926),Riedel (1935), and Stanley
(1928-1935). Roberts was the first to show that an increase in
temperature during the larval stage of vestigial flies increased the
length of the wing through all the intermediate stages up to a
normal wing. Like all his successors he expressed this change in
terms of wing length or growth. It is, in fact, something very
different, as will be shown later, but for the sake of description and
measurement these terms may pass for the present. The point of
importance at this juncture is that the stages of "lengthening" of
the wing reproduce the phenotypes of the series of allelomorphs of
vestigial. More exact data were obtained by Harnly (1930).
The following table gives his results in terms of wing length; it has

Table 2

Degrees

Mean length, mm.

centigrade

Male

Female

18.3

0.64

0.61

26

0.66

0.71

28

0.69

0.73

29

0.74

0.74

30

1.00

0.79

31

1.70

1.12

to be added that, with length, breadth also increases and that the
general shape of the wing (the amount of scalloping) assumes the

THE MUTATED Gh \ I 13

type of the ascending allelomorphic series named -trap, an tiered,
ragged, snipped, notched, and nicked.

It is important to note thai between 18 and 29 the temperature
effect was very small; a rise from 2<> to 30° affected the male wring
considerably; and from 30 to 31°, still more (three time- as
much)- In the female, the same happened bu1 began only 1°
higher. It oughl to be added thai general work in temperature
effects upon Drosophila proves thai 2!)° is aboul the upper limit
of the physiological temperature zone for this animal, beyond
which a rise in temperature has an effed upon developmental
processes thai in general terms may l>e described as inhibiting.
Stanley's work which led to the same results will he discussed in
the chapter on the sensitive period.

A still more detailed account has been given by Riedel (1935).
She also concludes that the wing of a vg-f\y becomes more and
more like Wild type with the rise of temperature, and she finds
definite shapes characteristic for each temperature. Figure 3
(( n pages 14 and 15) shows such a series with the temperatures
used. Their resemblance to an allelomorphic series is obvious.

In these experiments we are actually faeing the reciprocal of
the phenocopic effect, showing that at least in some cases the
process responsible for the development of the phenotype of a
mutation is reversible; furthermore, just as we found that the
phenocopies may be produced in a quantitative series imitating a
series of multiple allelomorphs, this reversed phenocopic action
also produces the phenotypes of the respective multiple allelo-
morphs of the ygf-series. \\ '<• shall return later to this very impor-
tant point.

C. The Sensitive Period

Half a century ago. the first workers in this field, studying
the temperature aberrations of butterflies, mainly with phylo-
genetic end- in view, made the important discovery that the
experiments were successful only if the stimulus was applied

during a rather short period of larval life preceding pupation.

This period was called the sensitive, or critical, period. All later
studies confirmed its existence and showed its importance
for an analysis of the problem. We shall describe the facts

Using the same examples as before with the addition of new
cases.

1 1

PHYSIOLOGICAL GENETICS

It baa already been stated I bai the production of phenocopies in
Drosophila succeeded only if the temperature shock was applied
at a definite time in development, the sensitive period. It could
be shown (Goldschmidt, 1935a) (1) That some of the phenotypes
could he produced only during a rather short period; e.g., the

Fig. 3a.
FlG. 3. — a, »0-wings at different temperatures; b, vg male wings at 32.9°C.
(1st col.) and at 34° (2d and 3d col.). (From Riedel, 1934, Arch. Entwicklg-
mech. 132, Figs. 12, 13.)

type with scalloped wings can be produced only during a short
period preceding pupation; (2) that other phenocopies may be
produced over a rather considerable stretch of time in develop-
ment, provided that a sufficient stimulus is provided; (3) that
the sensitive periods of different phenotypes may overlap; (4)
that the sensitive periods ceteris paribus are different in different
races. Table 3 gives the results for a few types in percentage
of obtained phenocopies in one experiment.

THE MUTATED OENE

15

10

PHYSIOLOGICAL GENETICS

Table 3

Phenotype of wings

Temperature shock applied at nth day of
larva at 25°

5

5K

5H

5H

6K

6H

QH

7

7'4

73A

Curly

79

20

8
4

29

67

80
4

100

100
50

33

58
4

SO

Spread

Scalloped

Lancet

Miniature

16

21

B I

i r

I

Jit

ra

||

I

II

II

11

1

1

1

|

ii 1

1

II11

II

1

1

1:

It is important to note that Friescn (1935), in producing;
phenocopies by X rays, found, as far as I can see, the same sensi-
tive periods. Thus he remarks
that the period for spread wings
occupies the last larval and first
pupal day, whereas scalloping
is produced one day earlier. It
might be added that very exact
determinations of the sensitive
period are difficult in this case
because the lethality after treat-
ment is so high that mass cultures
with a considerable variation in
time of development must be
5 used.

Far more exact determination
of the sensitive period is possible
in the cases of Bar and vestigial,
where the described effects are ob-
tained within a normal or almost
normal range of temperature. A
very considerable amount of work
has been done by Zeleny and his
school to locate exactly the sensi-
tive period for the temperature effect upon the Bar-eye series in
Drosophila. Figure 4 represents Driver's (1931) results for Bar
and Ultrabar.

In this connection, we are interested only in the fact that sensi-
tive periods exist and that they occupy a definite length in larval

600

540

480

i?420
o
-^360

' Q.300
"t

1240
c

i- 180
o

^ 120

60

0

30 27 2220 1715

Temperature in degrees C.

Fig. 4. — The main lines represent
the lengths of larval life at the con-
stant temperatures for Bar and
Ultrabar in Drosophila. The short
line to the left of each represents
the relative length and position of
the temperature-effective period for
the females; that to the right of
each, the effective period of the
males. (From Driver, 1931, J. Exp.
Zool. 59, Fig. 7.)

7 ///■; Ml TATED OENE 17

life. Another problem, which is illustrated in the diagram, is
whether this period is fixed as a definite stage in development or
its place in time is relatively different under different temperature
conditions. This special problem will have to be discussed later
in connection with the rate concept. In this ••use, the effective

period falls within the third instar, a time at which the eye

Aiihni! exists as an imagina] disk; according to Driver, it begins
about the time when optic and antenna! ArUagen are separating
from each ot her, i.< .. very early in eye development. It ends, as

expected, when actual facets are being formed.

31

30

29

J27

8 22

& 17
w

5 17
a.
E
<u

»- 30
27
22

17

Vestigial

lie terozygoles

Wild

0 50 100 150 200 250 300 350 400 450 500 550

Hours

Fig. ."). J):ita for the lengths of the developmental periods and tin- positions
and lengths of the temperature-effective period f>>r wing length in Drosophila.
The long linos represenl the earlier developmental peri ids. The solid portions
indicate the larval periods; the broken ones, the pupal periods. The upper short
lines represenl the temperature-effective periods for wing lengths of the females;
the lower lines, those of the males. (From Stanley, l'.».'5">. ./. Exp. Zool. 69,
Fig. 17.i

Another case upon which much work has been done is the
vestigia] wins. It was reported that the vestigial wing in
I >ro8opbila could be induced l>y increase of temperature to change
into the phenotype of all the other t^-allelomorphs. Stanley
(1935) determined the sensitive period for the production of
these phenocopies as well as for corresponding effects upon

18 PHYSIOLOGICAL GENETICS

t^-heterozygotes and also the small wing-lengthening effect
upon Wild-type wings. His results are summarized in the
graphic representation I Fig. 5).

[f these data are compared with the tacts known about the
development of wing buds, they point out that the important
phenocopic effed (30 and 31°) occurs about the time of the
third instar when the wing buds are just being formed. This is,
indeed, to be expected on the basis of the facts of development of
rowings, which will be reported later. There is, in addition,
another very early effect visible at lower temperature, an effect
that must have occurred at a time before the wing bud became
visible in the larva. In addition, there is a later sensitive period
for the heterozygote and the wild type. These discrepancies
are caused by a fact, unknown to Stanley, viz., that the effect
upon the vestigial wing is something different from the effect
upon the Wild-type wing and has nothing whatever to do with
growth in length. Only the 30 and 31° experiments upon the
vg-w'mg pertain to the phenocopic phenomenon and fix their
sensitive period. Harnly's (1936) determinations agree rather
well with Stanley's. He fixed the beginning of the sensitive
period for all the temperatures that have a phenocopic effect
(29 to 31°) at 64 to 68 hr. of larval development, which corre-
sponds to the molt between the second and third instar. The
duration of the period was, however, somewhat different at
different temperatures.

Let us interpret these facts. It is clear that the process con-
trolling the abnormal development of the wing buds sets in as
soon as their Anlagen are differentiated. From the facts con-
cerning the development of these mutant wings we shall later
derive the conclusion that the different degrees of wing scalloping
are dependent upon the time of visible onset of a degenerative
process. The sensitive period described here is, then, the period
in which this process may be counteracted, perhaps by changing
its relative velocity, the effects of which would be the same as a
later onset of the process. At first sight, it might seem to be a
contradiction that the sensitive period for the phenocopic action
in the Wild type occurs much later; the reason is that in the Wild
type, the temperature shock starts a process that otherwise
would not occur and the amount of which is determined by its
onset within a period in which such a process might still be

THE Mi TATED 01 VE 19

initiated before the wing structure is completely determined.
In the reciprocal effect, however, a process thai begins early in
development i- only -lowed down. The sensitive period then
covers such a time in development at which the process (early
degeneration caused by the vg-gene) has not yet gone so far that
it could not be Blowed dow ii more or less. The reciprocal produc-
tion of the same phenocopies, at least in part, from the Wild type

as well as from the mutant, is then actually the effect of two

completely different inductions, viz., in • case, the induction

of a process at a definite time and, in the other, the change of rate
of an already existing process; both are hound to lead to t he same

phenotype, because this is determined, given the same process

of degeneration, by the product of its rate of progress and the

time of onset.

Still more exact data are available for some of the bristle
mutations in Drosophila. Plunkett (1926) made such deter-
minations for the mutant Dichaete, and his student Child (1936)
for scute. These experiments do not exactly belong to this
chapter, as the effect of temperature here is not the production of
phenocopies. The scute mutant upon which the experiments
were performed makes certain bristles on thorax and scutellum
disappear. But the amount of suppression of development (or
destruction?) of these bristles varies considerably: a certain per-
centage of individuals shows the absence of one bristle and
correspondingly of all the bristles affected. This percentage is
influenced by temperature, and there is a temperature-effective
or sensitive period involved here. As the same effect may be
produced also by genetic modifiers, this case is at best closely
related to the case of phenocopies, and the underlying laws are
assumed to be the same. Child (1935) found that the tempera-
ture effects are more or less specific for each bristle but that the
sensitive period is the same for all, viz., situated at a time interval
from 75 to 98 per cent of the egg-larval stage. We recall that
in our phenocopic experiments the sensitive period tor bristles
in general (all effect- upon bristles) extended also into the pupal
stage. Plunkett also found such for the variation of the I lichaete
type. By still more refined experimentation and statistical

treatment of his data, Child was able to calculate the sensitive

period for the individual fly— as opposed to the determination for

a population with the following result: This period occupies

20

PHYSIOLOGICAL GENETICS

not less than 3.8 and not more than 7.") per cent of the egg-larval
period. The sensitive period for any one fly lies between the
time when 89.3 per cenl and the time when 96.8 per cent of the
egg-larval period have been completed. We shall meet these
tacts again in later chapters.

(In a recent paper, [ves (1935) criticizes the results of Child,
because ocellar bristles may be influenced at all times by a
temperature of 40 to 41°. But so high a temperature may act
very differently, by actual destruction of material.)

60

72

84

24 36 48

Pupal age in hours

Fig. 6. — Curves for the breadth of the area enclosed by the symmetrical bands
in the wing of Ephestia kuehniella after temperature action at different stages.
The horizontal lines represent the controls. {From Kuehn and Henke, 1936,
Ab. Ges. Wiss. Goettingen, Fig. 93.)

Finally, one case has to be reported which is of special sig-
nificance because it will furnish us later with important informa-
tion regarding the meaning of the phenomenon. In the extended
studies upon the wing pattern of the flour moth Ephestia kuehni-
ella (Zeller), Kuehn and Henke and their students furnished
many facts pertaining to this problem. They found (Kuehn and
Henke, 1936) mutants that produce a shift of the two systems
of symmetrical bands on the wings toward its center, thus nar-
rowing the central field, called the field of symmetry (see Fig. 7).
The same effect may be produced as a phcnocopy by treating the
pupae at a definite stage with heat shocks (45° for 45 min.), and
the sensitive period for this part of the wing pattern was fixed
at a time 48 to 60 hr. after pupation at 18°. Figure 6 represents

THE Mi TATED OENE

Central f<eld

21

Heat modification of width of central field in
percentage of wing breadth
24 28 32 36 40 44 48 52 56 60 64 68

Inhibition of

central field

after operation

Per cent animals

20 40 60 80 100

24 28 32 36 40 44 48 52 56 60 64 68 20 40 60 80 100

&

■ 11

I I f

Pi £fe ssfc

Sysy

st/st/ sysy

jybsyd

SybSydj Genotypes

Normal extent of 'central field

I ia 7. Influence of beal shocks of different genes and of then :auterinng

wings upon the breadth of the field of symmetry in the pupae of the Hour moth
Left series ot curves: influence of heal shocks at different times in a t >n >< »• 1
segregating into 2 Sy < Irdinates: percentage of individuals, abei

breadth ol area in percentage of entire \\ i 1 1 n . Right-hand curve: frequencj "t
incomplete formation of the same wing area after destruction of parts in per-
centage of animals operated at different agea. These ages correspond to those
for the curves on the left side. Below: type of wings belonging t" the different
points on the abscissa, The genes of the formulae produce the same effecl
genetically as otherwise produced by temperature shocks (phenocopii
Kuehn mill Henke, 1936, Abh Qet Wiu. Qoettingen, Fig. n»u

-22

I'HYSIOUUiUM. GENETICS

1 *-'

J

\

l

'"'

/ — i

y'"

\

/

\ \

\ l\
\ 1 \

a curve of the average diameters of this field on the wing after
heat treatment at different times as compared with the untreated
control-. A few further details regarding the amount of the
reaction may be gained from Fig. 7 to which we shall return later.
Similar determinations have been made for other elements of the
wing pattern in the same object as well as in Abraxas grosmlariata,
the currant moth. Some of the data will be reported later in
the chapter on pat tern.

$1Y

III

11

12 3 4 5 6 7

Da\/s

Fi<;. 8. — Superposition of a heat effect upon the dominigene effect in the
heterozygous zip- wing in Drosophila. Broken curve: amount of scalloping of
wings (classes I-V) in the controls, with genetic scalloping produced by the
^-dominigenes; average grade for individuals hatching on consecutive days.
Full-line curve: The same line treated with temperature shock: on the third
day the animals are hatching in which the genetic, and the temperature effect
are additive. {From Goldschmidt, 1935, Z. ind. Ahstl. 69, Fig. 17.)

It is not probable that in higher plants strictly comparable
facts will be met with, as developmental mechanics are so differ-
ent from animals. But Harder (1934) claims that flowers of
Petunia have a sensitive period in which the pattern may be
influenced by temperature and light in a definite way.

D. Combination Effect of Gene and Phenocopy

For our task — to prove that the processes underlying the forma-
tion of phenocopies are the same as those set in motion by
mutant genes — the following facts are of importance: Gold-
schmidt (1935a) treated with the method that produces scalloped
wings both Wild-type flies and flies that have genetically scalloped
wings (^-hcterozygotes with dominigenes). The result was that
in the latter case the genetic and the phenocopic effect were
added, and a more extreme scalloping was produced. Figure 8

THE Mi TATED OENE 23

represents the result of this experiment. The ordinate indicates
the degree of scalloping (classes I to V) ; the abscissa, the days of
hatching; and the dotted curve, the control. On the third day
appear the flies thai have been treated with heat, and the addition
effect becomes visible. Such additional effecl is, of course, to be
expected only in rather simple cases, it- occurrence depending
upon the detail- of the processes involved. Kuehn and Henke
(1930) state that in the alteration studied by them, i.e., in the
distance apart of the bands on the wing of the Hour moth, no
additive effect was found.

2. THE PHENOCOPIES AND THE DEVELOPMENT OF MUTANTS

Near the bottom of page 3 are pointed out the limitations
set to possible changes in developmental processes by the
necessity of their integration into a viable whole. These
limitations apply both to mutant changes and to phenocopic
changes. We saw thai these different causes produce the same
end effect, presumably by changing the same developmental
processes in an identical way. To understand, then, the type of
action of the mutated gene upon development we must be
acquainted with the differences between normal and mutant
development. 'Hie knowledge of this combined with the facts
already reported ought to give a first insight into our problem.

A. Facts of Mi iwi Development

To test these or other possible views it becomes imperative to
know the details of development that distinguish the mutant
types from the wild type. Studies of this type have been called
by Haecker phenogenetics, a term that is not accessary but is
sometime- useful. Quite a number of facts are known of various

importance for our problem.

A first group include- cases that allow only rather general con-
clusions. We mention the work of Pernitsch I 1913) and Schnak-
enbek (1921), who studied the development of dark and albinotic
axolotls. They found that both contain melanophores and
xanthophores and that in the albinotic form these pigment cells
have a lower rate of growth and multiplication. Goodrich 1 1927
found in the fish Oryzias a somewhat different behavior: In
different color varieties (mutants), the same pigment cells
received different quantities of pigment. Such cases, as far as

24 PHYSIOLOGICAL GENETICS

they go, may be interpreted <>u the basis of different rates of
reaction.

A second group includes cases that are rather complex in
detail, because the mutant form exhibits complicated patho-
logical conditions which are difficult to describe in simple terms,
although very thorough work has been done. This group con-
tains mosl of the recent work on mammals and birds. The
abnormalities arc caused by either recessive or dominant mutant
genes, the latter being sometimes lethal in homozygous condition.
There i< the work of Little and Bagg (1929), Bonnevie (1934), and
Plagens (1933) on a recessive mutation produced by X-ray
t nat ment by Little and Bagg. The abnormality is very diversi-
fied, affecting many different organs, especially eye and foot. It
seems that all these changes are the consequence of a blood
extravasate occurring in the neck of embryos 7 to 8 mm. in
length (Bonnevie). The blebs move with further development
to different places where they obstruct mechanically the processes
of development. We may not, in this case, speak properly of
changes in development produced by the mutant gene. It
actually produces only one effect: In young embryos, fluid from
the medullary tube is expelled through the foramen anterius in
abnormal quantity, and the blebs thus formed harm the develop-
mental processes wherever they become situated.

In some of the other abnormalities, however, actual develop-
mental processes are changed by the mutant gene. The develop-
ment of the short-tailed mutation of the mouse, found by
Dobrovolskaja (1928) as a dominant mutation, lethal in homozy-
gous state, has been studied by Chesley (1935). (The morphology
of the mutant is carefully analyzed by Kobozieff, 1936). The
homozygous embryos show, as far as can be determined, normal
development up to the somite state. Then a considerable
degeneration of tissue sets in, as indicated by the appearance of
chromatic granules, in the posterior part of the embryo. This
results in the absence of notochord, medulla, and somites and
leads to death of the germs (10?^ days after insemination). The
facts seem to point out that the degeneration process begins in
the primitive streak and therefore affects all the descendant
parts (and the induced differentiation) of this region. It is
obvious how much this case resembles the case of the vestigial
wing in its general features, which will be reported below. In

THE MUTATED GENE 25

the heterozygote, the developmental processes involved in pro-
ducing the short-tailed type are of the same order. I'm abnor-
mal development, affecting primarily the notochord, begins al
a much Later stage and therefore affects only (or mostly) the tail.
Here, then, the controlling factor is also the different time of
onsel of the degenerative process, as in the vestigial case.

The number of reports in this field of research is constantly
increasing, and the facts appear rather variable, though in general
they may be described in a similar way. Sere arc a few more
examples: van Assen (1930) finds thai genetic vertebral ankylosis

is based upon th currence of a fold in early development which

causes a shift in the time relations of the developmental processes.
I >avid (1932) studied hairlessness in different types and finds it to
be caused by very different processes in the different forms.
Sometimes development proceeds normally, but at a definite
point cysts are formed which de-troy the follicle. In still other
forms the follicle degenerates. Bonnevie (1936a) extended her
studies to the destruction of the labyrinth in Dunn's "shaker-
short" mice. In the development of these waltzing mice, the
labyrinth develops first normally up to the eleventh day. Only.
then does abnormal differentiation set in. It is caused by abnor-
mal development of the myelencephalon, combined with a
collapse of its ventricle. A consequence is that the auditory
vesicle is not innervated and doe- not continue normal differen-
tiation. We might finally mention in this connection the radium-
induced mutant abnormalities of Antirrhinum, which Stein (1932-
1935) produced. Here, also, a primary tissue destruction of a
cancerous type leads to formative changes and abnormalities,
though not strictly comparable to the case- in animal.-.

In the fowl, two comparable abnormalities have been studied.
There is the creeper fowl, showing a general achondroplasia of
the bone- in heterozygous condition, the homozygous mutant
being lethal il.andauer and Dunn. 1930//). The morphology
ami physiology of this mutant have been studied most thoroughly
by Landauer in a large series of paper-, and we know more about
the effects of the creeper gene than about any other comparable
case. We -hall report the case and other similar one- in another

chapter i -ee paije 214).

Another comparable abnormality, caused by a dominant factor

(viable when homozygous), is the rumple-- fowl, in which the

26 PHYSIOLOGICAL GENETICS

axial skeleton is shortened and ankylosed in the posterior region
(description in Du Toit, 1913; genetics in Dunn, 1926). The

details of development have not yet been worked out, but we have
already mentioned the fact that this condition has also been
produced as a phenocopy by Danforth (1932) (see page 10).

It is rather remarkable that the type of genie effect upon
development found in vertebrates may also occur in insects.
This means an early effect upon a more generalized develop-
mental process which leads to a number of later consequences.
Such facts were described by Speicher (1933) for the development
of the eye mutant Eyeless in the wasp Habrobracon. Here the
development proceeds regularly until a late larval stage. At this
time, a portion of one pair of imaginal disks fails to invaginate
sufficiently to become separated from the larval hypodermis.
This leads to a retardation of the development of the dorsal part
of the head capsule, resulting in the formation of large lobes in
that region. Later, all the elements of the eyes and head appear
in normal order but are modified in size and shape by the previous
event.

A third group contains work in Drosophila, where some decisive
facts could be found in the work of Schultz (1935) on eye colors,
of Medvedew (1935) on eye form, of Goldschmidt (1935c, 1937)
on the mutants of the wing, and of a few others. As the last
eases seem to be the simplest for explanation, and as here the
parallel with the phenocopies has been established, we shall
describe them first.

One of the phenocopies that may be most easily produced both
by heat and X rays are scalloped wings. Their type covers
exactly the phenotype of the vestigial allelomorphs nicked
(vgni), notched (vgno), up to snipped (ygsn). The sensitive period
is situated at the end of larval life. A study of the development
of the series of ygr-alleles gave remarkable results (Goldschmidt,
1935c, 1937). Phenotypically a wing of this series appears like a
complete wing from which parts have been removed starting at
the tip and proceeding toward the base of the wing until only a
small stump has been left in the alleles vestigial and No wing
(see Fig. 3). (This was known to the first observers but neglected
in later work.) The morphology of this series already indicates
that the typical subdivision of the wing surface with its veins
and spaces between the venation must have occurred in develop-

THE MUTATED OENE

-'7

1 1 n 'i it (if not \ isibly, .-it Least in the form of spatial determination)
before the process that results in scalloping begins. Actual
development proves that this expectation is literally true. In the
scalloped types, almost up to the type of the allele snipped,
development of the wing proceeds perfectly normally until
pupation, when the wing disk is protruded (evaginated) and
expanded to the size and shape <»t" the pupal wing, which secretes
its chitinous sheath (Fig. 9). Only then a degeneration, <>r

t

I

r

Fio. 9. Three stages From poatpupal development of a vg og** whig in 1 l
phila. The wing ia normal at the time of pupation. (From Ooldachmidt, 1937
of ' alif I , i>l. 15.)

lysis, of tin- tissue sets in at the point- of later scalloping, and
these pathological area- of the \\in«j spread are gradually resorbed
to form the nick- and notches (Fig. 9). This resorbing process
end- when the detail- of wing -tincture are being perfected.
It i- a general rule that as the process of degeneration sets in
earlier and earlier, a more and more extreme scalloping results.
In the higher made-, ,'., ., the phenotypes ragged, antlered, -trap,
vestigial, and No-wihg, the process actually begins in the imaginal

disk, and at the time of pupation part- of the wing area have

28

physiological a i:\etics

already been destroyed Figs. LO to 14; Goldschmidt, 1935c, 1937;
A.uerbach, 1036). In the highest types like vestigial, the
process must begin in the young imaginal disk. This, however,
cannot be demonstrated but has to be extrapolated from the rest
of the scries.

Similar results have been obtained for a number of other wing
mutants like dumpy, cut, pointed, miniature, and beadex, all of

• -

I:

'•■■ ■-'

■ ■

Wf! '

Jftil

Fig. 10. — Three stages from postpupal development of a Drosophila wing
heterozygous for vg/vgn°. At the time of pupation the wing is already incom-
plete. (From Goldschmidt, 1937, Univ. of Calif. Publ. Zool., pi. 16.)

which show normal development up to a certain point, followed
by an onset of degeneration of tissue at typical points (different
in different cases) or by a slowing up of certain growth phe-
nomena of the already differentiated wing. There are, however,
also cases where obviously a different growth rate is present from
the beginning (mutant expanded).

These facts, then, show that the mutant gene makes something
necessary for normal differentiation to be absent or below the
necessary threshold after a certain time in development (or, vice
versa, produces something acting as a lytic substance above a

THE Mi TATED GE VE

29

Fi<;. 11. — Three stages of development of a Btraplike Drosophila wing. Con-
siderable abnormality already at time of pupation. 'From Goldschmidt, 1937.
Univ. of Calif. Publ . Zbol., pi. 17. i

Pto. 12. Development of t In- op-wing in Drosophila after pupation. Most
of the wing area destroyed before and much after pupation. ■/■" vm Qoldachmidt,
Biol. Centrbl. 55 54.

30 PHYSIOLOGICAL GENETICS

certain threshold); and furthermore, thai the time of visible onsel
of this gene-controlled deficiency is a simple function within
the increasing series of allelomorphs. Without going into further
details here, these facts indicate that here, again, processes of
definite velocity are controlled by the different mutant genes and,
also, processes having velocities of different but typical rate in
series of alleles with increasing phenotypical effect.

It is encouraging that Harnly (1936), whose studies on the
sensitive period of the vestigial series have been mentioned,
derived from his facts (which have to be restated in different
terms, as the development of the vg-wing was not yet known to
him) a very similar interpretation. He concludes (1) that there
is a definite pattern of wing development in time (in a later
chapter, we shall see that it is not pattern of development but
pattern of a determination stream); (2) that the degree of expres-
sion is dependent upon the duration and rate of processes
occurring in the larval period of the vestigial genotype; (3) that
mutations at the vestigial locus apparently affect the duration
and rate of these developmental processes.

A considerable amount of work has been done on the eye
mutants of Drosophila, though the results are far from being
complete. The eyes are formed from an imaginal disk which
separates from the pharynx as the so-called cephalic complex
after 8 to 16 hr. of larval development (total 206 hr. at 27°C).
Between 36 and 48 hr. at 27° this complex is divided into optic
and antennal buds, the former now being the eye disk proper.
This grows and differentiates until it is finished — except for the1
pigment — at about 84 hr.

The mutants of eye form have been studied by Chen (1929),
Krafka (1924), Johannsen (1924), and others. The most com-
plete account thus far available is for the mutants conditioning
decreasing size of the eyes glass2, eyeless2, Lobec, studied by
Medvedew (1935). This author finds that the differences
between these types (including the variability and asymmetry
of eyeless) are already visible at 24 hr. (27°C), though they are
rather small. From this time on, growth proceeds perfectly
parallel in regard to percentual increase. Figure 13 shows the
curve of growth for these four types, age of larva plotted against
logarithms of disk size. The initial differences as well as the
parallel growth rate is visible. The drop of the curves at 48 hr.

THE MUTATED OENE

31

indicates that up to this time the cephalic complex was measured,
and later the eye disk alone. The actual differences are then
determined very early in development. Unfortunately, oothing
at all is known a- to how these differences originate whether
fewer cell <li\ isions take place or an .1 nlagi of normal sizeissecond-

1000
900
800
700
600
500
400

300

200

100

40

zo

24 36 4d 60 72 84 96
Pupae
Fig. 13. The logarithmic growth curves of the imagina] eye disks of Wild t >j .«• .
glass, eyeless, and Lobe larvae (between 36 and 48 hr. the cephalic complex
separates the eye <lisk propel - From Medvedev, 1936, /.. tad. Abstl. 70, Fig. 3.)

arily reduced. Thus at present the only conclusion to l>e drawn
i- that genes are involved which control early processes of
differentiation. It seems thai the story is a differenl one for
the Bar-eye mutants of Drosophila. According to Chen, the
youngesl stages of the Bar eye have the same aze as those of the
Full eye, and only later do the size differences appear. No details
are yet known; hut certain experimental fad page 76)

i

.\Wt

t

11 1

\ /

7

/

y

7

32 PHYSIOLOGICAL GENETICS

point to the possibility thai in this case a phenomenon may
occur similar to that observed iii the vestigia] case, viz., a second-
ary though rather early destruction of already formed eye
material.

A few facts, published recently by Wolsky and Huxley (1936),
actually point in this direction, though they are in no way con-
clusive as yet. In the Bar eye, there are small areas adjacent
anteriorly and sometimes posteriorly to the region of faceted
ommatidia in which no facets are found, though the pigment is
present. The authors found that in this region the retinula cells
and the dioptric apparatus arc lacking and only the pigment-
bearing accessory cells are present. This fact may or may not
permit conclusions upon the formation of the Bar eye.

A case of a different type is the development of the eye-color
mutant characters of Drosophila. Schultz (1935) has made a
special study of the eye pigments of Drosophila. It was known
from the work of Johannsen (1924) that the normal Wild eye
contains in its pigment cells deep red, orange, and yellow granules.
These are contained within the so-called primary and secondary
cells and beneath the basal membrane (see Hertweck, 1931).
The yellow granules are rather rare and occur only in the primary
cells. These pigments appear first about 50 hr. after pupation
(25°C), first around the basal membrane, then in the secondary,
and last in the primary cells. The general color at this time is
tan. At about 80 hr. the color changes to red, again in the same
centrifugal seriation, still leaving some yellow granules in the
primary cells.

Both the red and yellow pigments are water soluble and may be
separated chemically, and the red pigment can be oxydized into
the yellow one. Reciprocally, the tan pigment of the early pupa
may be reduced to red by action of H2S. This means that only
one pigment, in different states of reduction, is present. The
eye-color mutants, of which so many have been described, are
of two main types. One group shows bright reddish colors.
These eyes contain dense yellow or orange pigment in the cells and
little pigment below the basal membrane. The second group
of more brown shades contains hardly any pigment in the primary
cells; red and yellow or only yellow granules in the secondaries.
There is also an intermediate pinkish group of the pattern of the
first group but with red and yellow granules. Furthermore, in

THE MUTATED GENE

33

addition to the stage of oxidation, the pattern of pigment
distribution is involved as well as the amount of the two main
pigments present. Within the general frame of this grouping,
innumerable variations may be found in regard to both total
amount and relative concentrations of the pigment.

The study of development reveals that there is a characteristic
feature regarding the development of these colors. The follow-
ing table taken from Schultz shows the percentages of pupal
development completed at the time when pigment first appears
for different eye-color types.

Table 4
(From Schultz)

Type of eye color

Group I:

apricot. .

brown . . .

carnation

clot

prune

purple. . .

sepia ....

wild

Group II:

garnel . . .

light

peach

ruby ....
Group III:

cardinal

cinnabar .

scarlet . . .

vermilion

Percentage of pupal devel-
opment complete at

First appear-
ance of color

53
53

48
52
49
51
55
55

56
57
55
53

63
71
68

71

Darkening
of bristles

81
83
79
74
76
71
79
83

77
82
74
74

74

77
77
75

There is a first group behaving very much like the Wild type,
and a second similar one. But with the colors of the third group
the pigment appears only at the time when the Wild type changes
from tan to red (coinciding with darkening of bristles). There is,
then, a typical difference in the onset of pigment formation.

,;i PHYSIOLOGICAL GENETICS

As far as this work goes, it proves that the embryological differ-
ence between the different mutants consists in differences
(1) in time of onset of the process of pigmentation; (2) in
the amount (or rate?) of oxidation and reduction; and (3) in the
pattern of deposition of pigment (probably also in the size of the
granules). Such a combination of reactions of different rate
as found here had actually been postulated by Goldschmidt
(19206, 1927c) as a basis for the explanation of this and similar
cases.

As a supplement to the results on the Drosophila eye, the facts
that Wolsky and Huxley (1934, 1936), described for the develop-
ment of eye mutants in Gammarus are of importance. There
are mutants in eye structure in which the retinula is absent
or reduced to scattered cells, the crystalline cones are defective,
and the eye is composed mostly of hypertrophied connective
tissue. Furthermore, the optic tract is affected (which is also
the case in the Drosophila eye mutants). The development of
these eyes is abnormal from the outset: the retinula cells arc
deficient in number, do not form regular groups, and later
degenerate. Simultaneously the interstitial cells hypertrophy.
The centrifugal differentiation of the optic tract with its three
ganglia is delayed and inhibited. The adult tract then resembles
a developmental stage of the normal eye. As far as this descrip-
tion goes it seems that this development is comparable to the one
of the vestigial mutant; very early probably the Anlage is normal,
but early in development degeneration sets in, beginning in the
distal region of the organ. From these facts we may expect
that a histological study of the early stages of the Drosophila eye
mutants might reveal similar conditions and not simply a retarded
growth.

Some of the phases of the process of pigmentation as dependent
upon mutant genes seem to be more easily understood in birds.
The pigmentation of the feather and the general basis of feather
colors has been the subject of numerous researches of which we
mention especially the school of Haecker, because they worked
with genetical ends in mind (Haecker, 1918, 1925; Goernitz,
1923; Kniesche,' 1914; Spoettel, 1914). In a general way, the
different colors are produced by a combination of (1) Eumelanins
= dark relatively insoluble pigments ; (2) pheomelanins = yellow

THE MUTATED GENE 35

or brown soluble pigments both formed by oxidation of dioxy-
phenylalanin (Dopa) through an oxidase according to Bloch;
they are always deposited in the form of pigment granules of
different shapes; (3) The yellow lipochromes, derived from
carotins, stain the horny feather cells in solution. (4) With these
pigments the structural colors act to produce the phenotype.
(a) The Avails of the external cortical cells of the feather rami
contain fine air tubes which produce the blue color of cloudy
media. The blue is enhanced by a black background; the
addition of lipochrome color to both produces green; (b) If the
superficial layer of rami and radii splits into fine scales, another
optic effect is produced responsible for the bluish grey of pigeons,
etc. It is again blued by a background of black; (c) An iridescent
effect is produced if the radii containing melanin are broadened
(Rensch, 1925), and the thin horny scales above them act as thin
lamellae. A study of these colors as controlled by Mendelian
genes in the buderigar (Melopsittacus undulatus) has been made
by Steiner (1932). In the development of the feather germ,
pigment appears first in the so-called dendritic cells, i.e., branched
melanophores, which are situated at the basis of the future
branches which are formed as rows of cylindrical cells. In the
green race, which contains dark pigment (besides lipochrome
and the blue effect), simultaneously the cylindrical cells of the
young rami produce melanin. In another race with gray wings,
which contains less melanin, the dendritic cells contain more
melanin, and the cylinder cells less; and in a yellow or white race,
still more pigment is contained in the dendritic cells, and none
in the cylinder cells. These dendritic cells, however, do not
enter the formation of the feather but are later removed when
the young feather breaks through. In the three cases, then, the
same amount of pigment is formed but is distributed differently
upon the feather cells and the impermanent dendritic cells. This
may mean either of two things: (1) The chromogen is formed
in the dendritic cells and diffuses from these to the feather cells.
The time at which oxidase is released means pigment formation
and the end of diffusion. Earlier release, then, means less pig-
ment for the cylindric cells. (2) The dendritic Cells may reach
a threshold for pigment formation from chromogen at an earlier
time and thus absorb more of the available chromogen. These

36 PHYSIOLOGICAL GENETICS

two interpretations, proposed by Steiner, fall completely in line
with all the other facts discussed in this chapter, viz., a genie
action through the regulation of the velocity of one or another

process.

In the case of blue forms, the lipochrome is involved only in so
far as it is completely absent. If this deficiency is combined
with low melanin formation (as in yellows), the result is white.
The olive colors, however, depend upon a change in the arrange-
ment of the air-filled tubules in the medullar cells; the respective
gene has therefore a morphogenetic effect. (Regarding different
degrees of lipochromes, Duncker has derived physiological
explanations from his genetic work on canaries and buderigars.
But his analysis has not been substantiated by chemical or
morphogenetic facts.)

As a rather remarkable case might be mentioned the develop-
ment of the mutant aristapedia of Drosophila, according to
Balkashina (1929). This mutant belongs to the group of
homoeotic changes, where one organ is transformed into a homol-
ogous one. In this case, the bristle on the antenna, called
arista, is transformed into a tarsus by a simple recessive mutation
(Fig. 14). In normal flies, segmentation of the antenna in the
imaginal disk begins in a larva of 4 to 4}4 days (25°) and ends in
the young pupa. In an aristapedia fly, however, segmentation
begins in a 2-day-old larva, when the segmentation of the leg
Anlage also takes place. This segmentation is at once of the leg
type and continues thus. We shall return later to this remark-
able case and point out here only that the immense morpho-
genetic effect seems to be produced by nothing but an earlier
incidence of a morphogenetic process.

The cases thus far considered all involve changes in develop-
mental rates. The actual processes involved are not clearly
indicated, although it is possible to draw inferences about them.
There are also cases of a much simpler type, some of which will
be reported in the chapter on rate-controlling genes. Here only
two of them may be mentioned because they illustrate two
simple types of mutant development. Gabritschevsky and
Bridges (1928) studied the development of the giant race of
Drosophila and found that it is indistinguishable from normal
development up to the time at which pupation ought to occur.
The giant larva then continues to grow a few more days before it

THE MUTATED GENE

37

Fig. 14. — a, antennae of normal fly; b, antenna of aristapedia; c, last joint of
aristapedia compared with normal tarsus (right). (From Balkaschina, 1929,
Arch. Entwicklgmech. 115, Figs. 1, 6, 7.)

38 PHYSIOLOGICAL GENETICS

pupates. As we know now from the work of Buddenbrock
(1928-1931) and Wigglesworth (1934-1935) that pupation is
controlled by hormones, this might mean thai the mutant gene
causes this hormone to be produced at an abnormally late time.
A case that might he called the reciprocal of this has been studied
by Dobzhansky and Duncan (1933). In the Drosophila mutant
chubby (a larval shape), the abnormal growth ratio, leading
to the changed proportions of the larval body, has already
occurred during the embryonic stage and is completed when the
young larva hatches. But this is an example of rate genes which
will be treated in a later chapter.

As there is not much botanical material available regarding the
development of mutant characters, two investigations might be
mentioned here w hich show- that probably the same type of proc-
i iss< is will be frequently found to be involved. Anderson and Abbe
(1933) found that the considerable and manifold morphological
differences between Aquilegia vulgaris and the mutant comyacta
are altogether the consequence of an earlier onset of secondary
thickening of the cell walls in the development of the stems.

A somewhat more complicated and therefore less clear case,
the development of ear form in wheat, has been studied by Philip-
chenko (1929). The complication is derived from the fact that
quite a number of genes are involved in controlling the form so
that it is not actually a difference in regard to the action of one
mutant gene that has been studied, though an effort was made
to isolate individual actions. The results may be summarized
thus: there is a definite general pattern of development of the
ear, of its growth and differentiation of parts, in all hexaploid
wheats. The genes for form of the ear do not affect this general
development but act upon it in different ways of inhibition.
There is inhibition of differentiation of the spikelets (genes ex and
p) and inhibition of growth of ear (gene C) and its parts (gene N).
Some of these genes act relatively early: at the beginning of the
general differentiation of the ear (eip) or about midway (CiN).
But the majority of genes produce such changes only toward the
end of development.

The same general type of development occurs also in the tetra-
ploid species of Triticum but is different in T. monococcum and
in Secale, a point that is of more interest, however, for problems
of phylogeny.

THE MUTATED GENE

39

B. Mutant Character and Developmental Stage

In the case of the Gammarus eye, we have encountered the fact
that the adult mutant type shows a condition that may also be
described as the fixation of a definite developmental stage This
is clearly an important point for an understanding of the action
of the gene, as was emphasized by Goldschmidt in the first stages
of his work on interscxuality (19176, 1920c) and therefore some
more facts ought to be mentioned at this point of the analysis.

Fig. 15. — a, male genital armature of Lymantria dispar. Pe penis, Ps sheath
of penis, Ri segmental ring, Sa Saccus, U uncus, Va valva; b, the same in low
intersexuality, also after temperature treatment; uncus double. {From Gold-
schmidt, 1920.)

A very frequent type of abnormality is the Hemmungsmiss-
bildung. This means that development has simply stopped at a
definite stage with the result that an embryonic condition is
carried into the adult stage, with such modifications, of course, as
are impressed upon the organ by the further development of the
whole. These abnormalities very frequently involve incomplete
concrescence of embryonic parts, like harelip, hypospadia, cleft
eye, and cleft palate. Many such types are also caused by

40 rilYSlol.oCKAL <i EMETICS

mutant genes, the effect of which must be to stop further differ-
entiation at definite points and at a definite time or to retard it
so that it cannot be completed. The following example shows
such a process actually both in the form of a genetic effect and
in the form of a phenocopy (Fig. 15): In low-grade male intersexes
of bymantria dispar, the part of the genital armature called
uncus becomes paired, a transition from the unpaired male
condition to the paired female homologous structure (labia), and
this intermediate condition is caused genetically by a definite
genie combination. Kosminsky (1909, 1924, 1927) and later
Goldschmidt (19226) showed that this same condition may be
produced by action of cold upon young male pupae, and Bobroff
produced the same effect by X rays (1930). Kosminsky further
showed that the uncus in the male is first laid down as a paired
Anlage which later concresces; the treatment at the proper time
prevents concrescence just as the intersexual genetic condition
does. Another case of this type is the eye abnormality called
coloboma, consisting of incomplete coalescence of the eye bulb.
This occurs as an inherited anomaly and may also be produced in
rabbits by feeding turpentine to the pregnant mother.

The analysis of this and similar cases, a considerable number of
which are found in the different organs in case of intersexuality,
furnished many facts, which demonstrate that an embryonic or
developmental stage of an organ may become perpetuated in the
adult as a consequence of definite genetic conditions which
generally are of the same type as a mutation (here a mutant
condition of sex genes). We shall therefore not be surprised to
find that frequently a mutant type is identical with a develop-
mental stage of the Wild type. As examples, I mention the work
of Tammes (1934). She found that in Linum usitatissimum, a
number of genotypes may be isolated that are distinguished by
different grades of coloring of the seeds, in fact different steps
of the spreading of color around the seed. A study of the
development of coloration in normal seeds showed that all
the conditions found in the different mutant genotypes reappear
as stages of normal development in which color first appears at
the pointed end of the seeds and spreads along the edge to the
broad end of the seed. Both faces of the seed remain light.
Then this light spot is more and more restricted until the whole is
self-colored. Tammes also realized that the explanation of this

THE MUTATED GENE 41

case is to be found by applying the concept of rates of deposition
of pigment, which are changed by the mutant genes or stopped
before completion.

Another but probably somewhat more complicated case is
furnished by the different hereditary abnormalities of the human
hand, e.g., brachyphalangy. In normal human development,
the second row of phalanges is formed in the third month, but its
development is not finished until after that of the rest of the
bones of the hand. In brachyphalangy, they do not appear until
later and attain therefore only an embryonic condition if they
continue development at all. (For facts see Pol, 1921 ; Mohr and
Wriedt, 1918.) These cases will be cited in a later chapter on
rate mutations.

A complete survey of mutant types of animals and plants
might furnish innumerable such examples, which have been
described as abnormalities by arrest of development. One such
instance is to be found in the anomalies of the human eye as
analyzed by Waardenburg (1930). He analyzed thoroughly a
case of hereditary ptosis combined with different abnormalities
of the eyelids and lachrymal apparatus. By careful comparison
with the embryological facts, he came to the conclusion that in
this as well as in other hereditary abnormalities of the eye an
embryonic condition has been preserved as the result of a gene-
controlled disturbance in the interplay of developmental reac-
tions, of course followed by such adjustments as are forced upon
the organ by the further development of neighboring parts. (We
might point out here also the phylogenetic significance of such
facts. Waardenburg mentions the theory of Boek that human
morphology is a "fetalization" of the morphology of apes.) I
have quoted in this connection in a former book the fact that the
rudimentary wings of carabids represent the stage of pupal wings.
Such examples could be multiplied indefinitely.

There can then be no doubt that the action of mutant genes is
very frequently of this type. In dynamic terms, it might be
described as an insufficient production of substances needed for
growth and differentiation, an insufficiency that reaches at a
definite point of development a subliminal condition instead of
coming up to the necessary threshold value. The basis for such
a description of the underlying process is derived from the exist-
ence of series of such effects with different times of onset of the

42 I'HYSIOLOCICAL GENETICS

process dependent upon a series of multiple alleles. This,

however, anticipates the discussions of a following chapter.

C. Lethality

It is one of the elementary facts of genetics that mutants with
rather Large somatic effects tend to be lethal. In terms of gene
action, this would mean that the change in developmental
processes produced by the mutant gene is of such a magnitude
that the proper coordination and integration of the different but
simultaneous processes are put out of gear. The study of the
phenocopies together with certain genetic facts brings these facts
into line with the general concept of gene action.

In Goldschmidt's work on the vestigial wing in Drosophila, it
was found that the sensitive period for scalloping of the wings
occurred a short time before pupation. It was further found
that the maximal phenocopic effect produced at this time by heat
treatment corresponded to the effect of the alleles notched to
snipped or to grade V to VI, if the amount of scalloping is divided
into 10 classes with nicked as class I and vestigial as X. The
study of development of flies with different allelomorphs further
showed that no visible degeneration of wing area has taken place
at the time of pupation, if the maximal final result is class VI of
scalloping. This means that the mutant genes up to snipped will
act in such a way that the visible effect does not occur before the
sensitive period and presumably that their action occurs actually
within this sensitive period. We know, furthermore, that the
higher alleles produce a similar visible effect before this period
and the highest alleles even in early larval stages. We might
infer from this that other sensitive periods exist in which definite
gene action takes place, maybe before each molt, and the work of
Harnly points to such a period at the molt from the second to the
third instar.

It is now knowm that the imaginal disk which produces the
wing also contains the material for the thorax segment and that
this material becomes separated from the wing material only
in the course of development. A degeneration of tissue produced
by the ^-allelomorphs will therefore affect the development of
the thorax segment also, if the action takes place very early. It
seems that the onset of the action of the y<7-gene occurs just near
this point in development; in some lines of vg, individuals with a

THE MUTATED GENE 43

half thorax are exceptionally frequent. (Occasionally they occur
also in long-winged stock.) Such animals are at the limit of
viability; if one side only is affected, they may survive; both sides
affected would be fatal.

Here, then, we are facing the point where a gene effect begins
to be lethal, and we see why: The action of the gene comes into
effect so early in development that a group of cells is affected
that still has a rather unrestricted prospective potency, which
would be restricted or segregated only in later development.
The effect of the gene-controlled process therefore influences a
whole group of processes of differentiation instead of a single one
in case of an effect at a later stage at which the material is already
subdivided in regard to determination. Therefore these early
effects tend to be lethal. In the v^-series, this is especially clear:
the highest alleles of the No- wing type are lethal in the homo-
zygous condition. Probably comparable cases are more frequent
than is at present known.

The same facts are found in the cases of mammalian abnormali-
ties which were described on page 24. Here it was possible, at
least in one case, to compare homozygous and heterozygous
action. It was clear that the lethality of the homozygous effect
was a consequence of the early action of the gene in destroying
Anlagen of a considerable prospective potency, involving whole
organ systems.

The facts here described are of considerable general impor-
tance; they show that the action of mutant genes may be con-
fined to definite periods of development; this would not imply
necessarily that these genes are inactive otherwise. It would
probably mean that the tissues were in a condition to react only
at definite moments. And it seems, furthermore, that these
moments of susceptibility are those in which processes of deter-
mination take place, i.e., of subdivision or segregation of potencies.
If further work should prove this to be a rule, an important link
between genetics and development would have been established.

We do not mean to express the opinion that all lethal actions
are of this type. Certainly lethal gene actions may also occur
if some physiological process of vital importance is damaged or
completely inhibited. The best examples of this are the chlorotic
mutants of plants, which cannot survive for lack of chlorophyll.
But shoots of this constitution may grow upon a green stalk.

44 PHYSIOLOGICAL GENETICS

Similarly, merogonic tissue in Amphibia dies, except when grafted
or grown upon a normal host (Baltzer, 1933). The same applies
to haploid Drosophila tissue (Bridges, 1930). Otherwise
doomed tissues of a lethal brachyuric mouse may be grown in
vitro (Ephrussi, 1935), and lethal deficiencies may exist as mosaic
spots (Ephrussi, 1934). All these cases show a physiological
damage produced by mutant genes or lack of chromosome mate-
rial, which spreads within the tissues and may therefore be
counteracted by the presence of normal tissue. These facts do
not, then, belong to the problems discussed in this chapter.
The same applies to genie interactions which might produce
lethal effects by lack of some coordination. Kosswig (19296)
and also Goodrich (1935) showed that in hybrids of the fishes
Platypoecilus and Xiphophorus the pigment cells produce such
an excess of pigment that actually fatal melanotic tumors occur.
A similar effect in early development would, of course, also be
lethal.

D. Preliminary Conclusions

The study of the phenocopies, the sensitive periods, and the
development of mutants already permits some conclusions which
have been partly anticipated at different points of the description.
Two different sets of facts are pertinent. First, we consider the
cases where the phenotypic expression of a mutant trait is a
function of temperature within the physiological limits of adapta-
tion to temperatures (in Drosophila ca. 16 to 30°). In these
cases, e.g., Bar eye in Drosophila, the effect follows the well-
known rules for temperature action upon life processes. In this
connection, it is of no importance whether this effect is described
in terms of the Van't Hoff or the Arrhenius or any other formula.
The fact is decisive that the effect is of the same type as observed
in certain chemical reactions, without any additional features,
roughly described in the terms of the Van't Hoff formula as an
increase of two to three times for 10° rise of temperature within
the physiological temperature zone. In these cases, some of the
multiple alleles of the mutant produce an effect at normal tem-
perature that is situated within the range of the experimentally
varied effect at different temperatures. If, then, the effect of
raising the temperature is to increase the velocity of some reac-
tions or processes, it seems that the mutant (multiple) allelomorph

THE MUTATED GENE 45

exercises the same effect. Zeleny (1923) who first analyzed such
cases, as reported, summarized the results with the words:
"... demonstration ... of the fact that the gene Ultrabar
has the same type of reaction as a temperature difference. It is
possible to state the effectiveness of particular germinal factors
in terms of corresponding effects of temperature." Rates of
reaction were not mentioned in this connection, as Goldschmidt
had already done in 19206 in trying to explain the phenocopic
phenomenon (as well as the general action of genes), but no other
way seems available to describe these temperature effects and
simultaneously the effect of the mutant alleles. Such was done
by Goldschmidt (1927c) in regard to the facts found by Zeleny;
and in his last reference to his work, Zeleny (1933) himself writes:

Since the circumstances connected with the production of the Bar
mutants make it possible to formulate a theory of difference between
the genes of the series in terms of structural change, quantitative in
some cases, and since the resulting somatic effects may be measured in
terms of temperature effect upon an individual with unchanged genes,
it seems worth while to postulate, for purposes of experimental test,
theories of gene action in terms of relative rates of the physiological
effects they produce.

The second type of experiments did not use external agencies
within the range of physiological conditions but actual shocks,
which proved fatal if prolonged. It is characteristic for this
set of experiments that similar effects are produced by different
methods. Thus heat, frost, X rays, narcotics, C02 (the latter
in von Linden's experiments on butterflies) produced similar
effects. True, they were not entirely identical. Thus Friesen's
X-ray phenocopies in Drosophila were not wholly — though in
part — identical with those produced by heat; in the experiments
of Kuehn and his pupils, heat and frost had somewhat different
effects on certain parts of the wing pattern of the flour moth.
But in a general way, the same effect was produced. From
such facts, the early experimenters upon the butterfly wing,
who were not yet interested in genetical problems, concluded
that the effect was produced by reducing differentially the
velocity of some of the processes involved in the production of
the normal phenotype (von Linden, 1904). Goldschmidt (19206)
when analyzing the typical effects of temperature upon the
lopidopteran wing went a stop further: He stated that evidently

46 I'UYSIOLOGICAL GENETICS

different processes with a different temperature coefficient were
involved which therefore were influenced differentially (speeded
or slowed) by the experimental temperature. The same inter-
pretation of the facts was later used by Zeleny (1923, 1933) and
by Harnly (1935). I think, however, that this interpretation
fits only the results of experiments within the range of physiologi-
cal temperatures. Outside this range, temperature shocks will
act always by differentially slowing down some processes.

All the recent investigations bear out the essential correctness
of such a conclusion. Applied to the mutant gene, which acts
with identical effect as those shocks, this means that the mutant
gene acts by changing the rate of definite developmental reactions
or processes. Based on this conclusion, the further inference is
drawn that each gene acts through the control of rates of reaction
(Goldschmidt, 19176, d, 19206, 1927c).

It seems that this conclusion, which the author has always been
inclined to consider the basic insight into the action of the gene
and which originally was derived from experimental facts reviewed
on page 52, is being more and more verified in recent work.
Many of the authors who have performed experiments bearing
on this question have now accepted this viewpoint: Plunkett
and Child, after a study of bristles in Drosophila; Zeleny, in his
experimentation on Bar eye; Harnly, who studied vestigial wang;
Dunn, after a study of minutes in Drosophila; Huxley and Ford,
studying eyes of Gammarus; and many others, analyzing different
types of cases (Wright, Honing, Tammes, Sinnott, etc.). We
shall refer to these facts at their proper place in our discussion.

E. Some Pertinent Facts from Experimental Embryology

At this point, some facts ought to be mentioned which in a
general way supplement the conclusions thus far derived. We
have already described Danforth's experiments, producing the
rumpless type of fowl as a phenocopy by experimenting upon the
embryo. In this case, the type produced corresponded to a
well-known type produced by a mutant gene. Many similar
cases are known, in which the treatment produces definite
abnormalities, definite types of monsters, though not exactly of
the same type as known mutants.

Many experiments haVe been performed upon lower vertebrates
which resulted in the production of typical monsters. In fact,

THE MUTATED GENE 47

the whole set of recent experimentation upon amphibia might be
quoted. But what is important for our present discussion is the
production of monsters identical with those found in nature and of
essentially the same type as those caused by mutant genes.
We shall therefore not report upon experiments of transplantation
or removal of embryonic parts, which lead to monstrosities,
as these are not comparable to the effect of gene action, though
analyzing directly the localization of decisive materials (Lewis,
Spemann, Harrison, Holtfreter, etc.). Stockard (1913) was
the first to produce typical monsters, especially cyclopia, by the
action of magnesium salts upon early embryonic stages of
Fundulus. Other authors as well as Stockard himself found later
that the same effect could be brought about by many agencies
such as anesthetics (McClendon), acetone and butyric acid
(Werber, 1916), low temperature (Kellicott, 1916; Stockard,
1913), ultraviolet rays (Hinrichs, 1925), and X-rays (Little and
Bagg, 1929). This reminds us of the early experiments with the
butterfly wing, where about the same agencies proved effective
in changing the pattern. And the comparison is still closer if
we add that in the experiments just reported there is also a
sensitive period, viz., from the fertilized egg up to the stage of
neurula. It is therefore not surprising to see that Stockard
(1921) derived the same explanation, viz., an action of all these
factors upon the speed of certain developmental processes.
He states that all methods employed have more or less the same
effect, without apparent specificity. (Notice the parallelism with
the phenocopies.) The reason is that all these methods primarily
slow up the differentiation of the embryo. The state at which
this happens decides which type of monster will be produced.
Not all stages of development react, however, in the same way;
e.g., the treatment of fish eggs before gastrulation produces mostly
double monsters; after gastrulation, abnormalities of the eye.
Other periods of development seem to be inaccessible to such
experiments. From this it follows that there are definite sensitive
periods for this type of treatment, and they are found near the
period of visible origin of the organ in question. Stockard
thinks that the actual basis of the phenomenon of sensitive1
periods is to be found in the assumption that the organ at this
time grows and differentiates faster than the others. A retarda-
tion at this moment, therefore, has a differential effect. If we

is

PUYSlol.tiCK \L GENETICS

Leave the lasl Bpecific consideration out of account, it is obvious
thai these tacts and their interpretation fall completely in line

with the result- of the analysis of the phenocopies.

There is another interpretation of these results by Child (1925)
and his student Hinriehs (1925) which seems at first sight to be
different. According to Child, pattern is determined by a
metabolic gradient which begins with a dominating high level

Fig. 16. — Semidiagrammatic representation of the main series of grades of
otocephaly in the guinea pig. {From Wright, 1934, Genet. 19, Fig. 1.)

spreading to adjacent regions of a lower level. In general, the
anterior end of an embryo is supposed to be the highest level.
According to Child, high levels are more susceptible to injuries
than lower levels, and therefore the reported experiments are
bound to injure mostly the anterior end of the embryo, which is
supposed to control the pattern. This interpretation (which, by
the way, does not agree with the results of experimentation by
the Spemann school) is, however, only a specific way of express-
ing in terms of metabolic level the same thing that Stockard
expressed in more general terms of inhibition and velocities.

THE MUTATED GENE 49

The first step to link these facts with the problems that we are
studying here was taken by Newman (1908-1917) when he found
that the same type of monsters could be produced by crossing
widely distant forms of teleosts. Here, then, a genetic agency was
found to have a parallel effect, which now brings this set of facts
close to the problem of the phenocopies. Here belongs also a
finding by Montalcnti (1933), who reported numerous abnormali-
ties produced by crossing species of toads all of which may be
attributed to the effect of asynchrony of developmental rhythms.

The foregoing interpretations are still better supported by
cases in which a single Mendelian gene produces effects very much
like the ones described; there is the case of Little and Bagg (1924),
of monsters in mice supposed to be inherited in a simple Mendelian
way; another somewhat different one by the same authors (1929) ;
a case of Wriedt and Mohr (1928) of such a type in cattle. Also,
Little and Bagg in analyzing the abnormalities produced in
X-rayed rats reach conclusions of a similar type. They find
that the defects are in most cases the consequence of extravasates,
produced at different times, which interfere with the normal rate
of differentiation of the respective organs. In some respects,
the best analyzed case, though not so simple on the genetic side
and not studied embryologieally, is the one found by Wright,
the case of otocephaly in guinea pigs (Wright, 1934c; Wright
and Wagner, 1934; Wright and Eaton, 1933). This heritable
abnormality consists in a graded series of changes at the anterior
end, beginning with a reduction of the lower jaw. (2) Then
the ears are joined below the jaw by naked skin; (3) then there is
only a single ear opening on the throat; (4) the mouth and upper
incisors are lost; (5) the nostrils fuse; (6) the eyes become fused
more or less below the proboscis; (7) and the fusion becomes
complete. (8) Then the proboscis is lost; (9) the eye also; (10)
also the ear opening; (11) and finally no head is left except a small
median ear. Figure 16 illustrates these grades.

This type of monstrosity is inherited, though not in a simple
manner, a fact that does not concern us here, however, except
that we are not dealing with the effect of one mutant gene but
with a combined effect of, perhaps, two. There is no comparable
phenocopy in mammals, but, as we have already reported, in lower
vertebrates monsters have been produced by experimental
treatment of early embryonic stages which closely resemble these.

50 PHYSIOLOGICAL GENETICS

Wright's analysis of the facts led him to the following conclu-
sions: In accordance with the experiments of Stockard, etc.,
which were reported above, it seems that the monstrosity is
produced by an inhibitory action upon early stages of develop-
ment prior to the visible formation of the organ rudiments.
(This is, of course, found to be the case in all experiments on
determination.) But whereas the experimental treatment pro-
duces more diffuse and general effects, the genes in question have
a specialized effect of the type described as otocephaly. This
means, according to Wright, that the gene effects are relatively
direct and precise in their moments of action. In detail, Wright
favors Child's point of view : Any agent that inhibits the metabolic
processes of the early embryo before the primary simple gradient
pattern has become complicated affects especially the most
active region with consequences mostly upon the anterior end
of the medullary plate.

Other explanations of the embryological facts are, however,
possible. Lehmann (1936) concludes from his experiments that
the primary cause for the formation of these monsters is probably
a defect in the head organizer. This sounds, indeed, more
reasonable than the idea of gradients, which has not much of a
factual basis.

Reference has been made to the work of Little and Bagg
in which it was shown that abnormalities in mice caused by a
recessive gene and appearing phenotypically rather diversified as
clubfoot, Polydactyly, anophthalmia, etc., were the result of
blood extravasates formed in early embryonic life. A simple
embryological process then produced all those complex features.
A detailed analysis of all the defects of skull, mandibles, brain,
etc., combined in the picture of otocephaly showed that they all
may be traced to a small number of centers of abnormality:
ventral mandibular arch, olfactory placodes, cerebral vesicles,
median optic rudiment, and some others. "These in turn may
be related by the hypothesis that the basic factor in this class of
abnormalities is inhibition of the anterior medullary plate and
associated ectodermal placodes. ..." It may be that a
still earlier and more generalized stage is the critical one (see
preceding paragraph). The important point remains that an
early inhibition at a definite point of embryonic development is
responsible for the whole chain of later events. All the types of

THE MUTATED GENE 51

monsters, then, may be explained on the basis of different degrees
of inhibition, acting upon different stages in development and
lasting different times. The action of the mutant genes is
conceived, as mentioned above, as a general inhibition, to which
the regions most active at a particular moment are the most
susceptible. It might be debatable whether this special inter-
pretation is correct (a description in terms of organ-forming stuffs
or organizers or determination stream could easily replace the
description in terms of gradients; (see Lehmann)). But the
general conclusion is certainly correct, which Wright formulates
thus: "The case illustrates the extent to which morphological
changes of an apparently qualitative sort may be brought about
as a result of ramifying correlative effects from initial gene effects
of a simple quantitative sort." We shall realize when discussing
the formation of pattern that this result furnishes another
example for the general correctness of Goldschmidt's views on
primary pattern formation as controlled by genes.

The facts described in this chapter may be linked with the
contents of the next chapter by some facts discovered recently by
Bonne vie (19366). She studied the development of still another
hereditary abnormality in mice, pseudencephaly, caused by a
recessive gene and consisting of a complete upset of the arrange-
ment of the brain which rides like a wig on top of the head. Again,
it was found that the primary cause of the defect occurs very
early in development. The decisive point is that the develop-
ment of the primary head mesoderm is affected. It seems that
all individual developmental processes are perfectly normal but
that the relative velocities of growth have been changed: the
ectodermal head parts are growing while their mesenchyme
surroundings are still underdeveloped, which forces the former to
fold and to behave abnormally. The rest of the abnormal
development is the mechanical consequence of this primary
situation. In other words, the effect of the mutation is to destroy
the proper timing of two major interdependent developmental
processes by changing the rate of only one of them.

3. THE MUTANT GENE AND THE RATE CONCEPT

The facts reported thus far all point in one direction : they show
that the mutant gene produces its effect, the difference from the
wild type, by changing the rates of partial processes of develop-

52 /'// YSIOLOGICAL GENET h '.s

ment. These might be rates of growth or differentiation, rates of
product ton of stuffs accessary for differentiation, rates of read ions
leading to definite physical or chemical situations at definitr
times of development, rates of those processes which are respon-
sible for segregating the embryonic potencies at definite times.
Thus far such conclusions were derived only from the facts
regarding the phenocopies and similar phenomena. Originally,
however, the interpretation of gene action in terms of rates of
reaction was derived from a different type of facts, analyzed by
Goldschmidt (19176, c, 19206), viz., observed differences in rates
of certain gene-controlled processes. For a special case, the
pigmentation of rabbits, Wright (1916) had also derived the con-
cept of rate of production of an enzyme to explain the action of
different alleles for coat color (see page 73).

A. Rate Genes

Goldschmidt (1917cZ) first drew attention to rate genes and
their importance for an understanding of gene action. He
described different genetic races of the gypsy moth, Lymantria
dis-par, the caterpillars of which were different in regard to their
markings. The main difference was that in some races a pattern
of light markings on the back persisted throughout development;
in others, it became covered during development by a dark cuticu-
lar pigment which was deposited at a certain rate with the
progress of development. This rate was different in different
races and intermediate in heterozygotes. Thus different genetic
constitutions could be linked with different rates of production of
pigment during development. Numerous curves for the process
are found in Goldschmidt (19246).

Simultaneously Goldschmidt studied another phenomenon
which led to an explanation of gene action via rates of develop-
mental processes. The phenomenon of zygotic intersexuality
produced by crossing the sex races of Lymantria dispar permitted
such an analysis. It was shown that a definite event in the
development of intersexes, the turning point at which sex
changes, could occur at different times, depending in an orderly
way upon the genetic constitution of the individual. It was
argued that such a perfectly orderly dependance of the time of
such a change from a similarly orderly set of conditions of the sex
genes could be understood only if two competing sets of reactions

THE M I TA TED GENE 53

of orderly but varying velocities were involved. Sex genes, then,
were conceived as controlling the rate of production of sex-
determining stuffs. From these two sets of facts and their
genetic and embryological analysis, a generalization was derived
applying the principle of rate to all genie actions. (For details
see Goldschmidt, 19206,c, 1927c.)

Many cases have since been studied which show a connection
between mutant genes and rates of processes. In this chapter
only certain cases will be mentioned in which the rate concept
has not been inferred from the facts but where actual rates have
been measured. Ford and Huxley (1927) were able to measure
the rate of pigmentation of the eye of Gammarus chevreuxi. In
the normal dominant Black-eyed form, the eye pigment first
appears scarlet. It later darkens through brown and chocolate
into black. In the Red-eyed mutation, the darkening of the pig-
ment takes place much more slowly and also starts later. The
end of the process, therefore, reaches a stable phase before black-
ening is completed. These two different rates have been plotted
as curves and have been found to depend to a certain extent
upon external conditions. Genetically the genes involved con-
trol either the rate or the onset of the pigmentation process.

Very nearly related to the facts concerning pigmentation in
Lymantria caterpillars are those found by Goodrich and Hansen
(1931) for pigmentation in fishes. They studied the increase
in the number of pigment cells in brown and transparent goldfish
in the course of development (genes T and 7\). Figure 17 shows
the rate curves for the two homozygous and the heterozygous
forms which resemble closely Goldschmidt's curves for pigmenta-
tion of Lymantria. In this category would probably also belong
the results of Hashimoto (1936) with the silkworm if an embryo-
logical study had been made. There is a mutant gene in this
form causing an extra pair of abdominal feet. In heterozygous
condition, only the feet are produced; in homozygous condition,
the next segment is also influenced and has an extra semilunar
spot. In this connection, such gene actions might also be
mentioned as the genes controlling later or earlier growth of
feathers in the chicken according to Serebrowsky and Warren.

A whole group of cases is known in which the mutant gene
affects definite growth rates. Of course, in this case, one might
claim that there is no difference between the action of a gene

.-.I

PHYSIOLOGICAL 0 E N E TK 'S

controlling color or shape or :my other character and a gene con-
trolling growth rate. This is true. But as it seems that all
gene actions may be in the end reduced to the control of rates,
such cases where actually visible rates are involved are of a special
significance. Only a few typical examples may be given, which
illustrate the type of processes involved. There are some cases
in which the mutant gene changes the rate of differentiation
without any visible changes of growth. Thus the vestigial gene

450

400

350
I 300

D.
O

J 250

£
*S 200

E 150

100

50 -

\tt

15 20 25
Length in mm.

Fig. 17. — Graph of average number of melanophores in the three types of
goldfish, Common TT, heterozygous TTl, and transparent TlTl. (After Good-
rich and Hansen, 1931, J. Exp. Zool. 59, Fig. 7.)

retards development during the larval period by 5} ^ hr. according
to Harnly (1929). More typical are the cases in which the
mutant gene controls a different rate of growth processes.
Studies of such cases have revealed a few facts of primary impor-
tance. Castle and Gregory (1929) showed that the differences
between large and small races of rabbits, which genetically are
caused by multiple genes, are, in fact, differences of rate of cell
division, which begin early in development. Thus the growth

THE MUTATED GENE

55

curve, though starting with eggs of the same size, becomes
different from the onset of development.

Exactly the same was found by Goldschmidt (1933a) for the
differences in size of geographic races of Lymantria disjpar.
Figure 18 represents curves of growth for three such races, female

2000

1800

1600

1400

1200

1000

800

600

400

200

° Hich'moe

1 Central

/J Mas sac h

• o* ' Hachiman

/ SH<? 's* J Central
4H<?^s , * </ tiichinoe

E

m

nr

Pu

Fig. 18. — Growth curves plotting larval instars against weight in milligrams
for caterpillars of Lymantria dispar of different size-races. (From Goldschmidt,
1933, Arch. Entwicklgmech. 130, Fig. 16.)

and male, with weight plotted against the larval instars. The
differential growth rates become visible as soon as they are
measurable.

As an example from plants, the observations of Muntzing
(1932) on Galeopsis tetrahit may be mentioned. Here flower
color and speed of development are correlated, the white-flowered
plants beginning to flower later than the red ones. It was shown

56 PHYSIOLOGICAL GENETICS

that this phenomenon was due to the color genes involved, which
might control simultaneously the rate of cell division. In this
case, it could also be demonstrated that the respective speeds
were proportional to the number of red factors present. This
relation is especially impressive, since the difference occurs between
white-flowering plants having or not having the red gene.

There is another characteristic feature of such cases as these:
Frequently an influence of the cytoplasm upon such growth rates
has been demonstrated. It came to light in both Castle's and
Goldschmidt's work, and a considerable part of the work of
Wettstein and his students on cytoplasmic influence in heredity
(his plasmon) deals with structural differences dependent upon
growth rates. (The rate of division of the protonema cells in
different species of mosses is a genetic character, probably depend-
ent upon one or two Mendelian genes.) This is, after all, not so
surprising, as all rates of reaction are dependent upon the sub-
stratum in which the reaction takes place (see page 263).

B. General Experimental Evidence

In the maj ority of cases that have been studied experimentally,
the conclusion that the respective gene actions are of the nature
of a control of rates has been deduced from certain regularities
which could be described in the simplest form in terms of rates.
It might be said that almost all careful investigations in this field
have corroborated the general concept that genes act by control-
ling rates of processes or reactions responsible for orderly develop-
ment. The general mode of attack has been quite different in
different cases, which in part have already been mentioned in the
foregoing chapters.

1. A Typical Case. — We begin with the case of the vestigial
alleles and phenocopies, which was reported without inter-
pretation, because in this case, genetics, development, and
phenocopic effect are known. The genetics of the case — leaving
aside for later consideration the interesting dominance conditions
— give evidence of a remarkable quantitative order of visible
effects within a series of multiple alleles, their heterozygotes, and
compounds. Mohr (1932) has made a thorough investigation
of the allelomorphic series Wild type — nicked — notched — vestig-
ial— No-wing. Figure 19 shows the average phenotype of the
different homozygotes, heterozygotes, and compounds as they

THE MUTATED GENE

57

S*

5+ ^ §5 «!

58 PHYSIOLOGICAL GENETICS

appeared in Mohr's work. Above the pictures the constitution
is marked, and below the pictures the percentage of individuals
showing the phenotype (percentage of penetrance). We realize
that the normal wing of the wild type ( + "'/ + ''',) occurs also in
the heterozygotes between wild and notched (vgno), wild and
nicked (vg"'), and wild and vestigial (vg) and in the homozygous
nicked fly. The compound nicked-notched shows that 0.2 per
cent flies with a nick and 100 per cent scalloped flies are obtained
only in the compound vestigial-notched; all further data may be
read from the figure.

Here, then, a series of alleles (into which, also, the heterozygotes
and compounds fit) produce a perfectly orderly series of effects,
with a simple quantitative arrangement of the phenotypes
resulting. It begins with no visible effect and leads through the
steps of a small effect in few, in many, in all individuals to
increased and still more increased effect up to the maximum that
is physiologically possible, and this maximum is at the threshold
of lethality.

The study by Goldschmidt (1935c) of the development of these
quantitatively different grades of scalloping of the wings showed,
as reported above, that development is normal up to a certain
moment, when a degeneration of the tissue of the wing margin
sets in and proceeds up to a time in development when no
further destruction of wing area seems to be possible, and from
then on the rest of the wing finishes its normal development. It
was further showrn that, so far as the evidence goes, the earlier
and earlier onset of the degenerative process produces a greater
and greater amount of wing destruction. Thus the different
genetic constitutions are linked with a process that contains a
definite time element. Taking into account further the pheno-
copic effect (see page 9) of the same type produced if the
temperature shock or X-ray shock is applied at the proper time,
(as well as the facts to be reported later on a parallel shift of the
phenotype by dominance modifying genes), we may form a
picture of the whole process in terms of rates. At the present
juncture in the analysis of the case, this picture might be differ-
ently conceived in its details; but in its general form it must be
of the following type: The process of degeneration of the wing
tissue which results in scalloping may be due to the presence of
something that prevents — physically (permeability, viscosity) or

THE MUTATED GENE

59

chemically (poison) — further differentiation or, by lysis, destroys
existing tissue. Or it may be the consequence of the lack of
something necessary for differentiation, an insufficiency, e.g., of a
growth substance, a vitamin. If we use first the latter concept
for the sake of simplicity, we could represent the case in the form
of the diagram (Fig. 20). The series of wj-genes controls the
production of the growth substance in question which has to
be present at each stage of development in sufficient quantity to

W

quant, wild rypejimitam^

10///
9//I

Nicked

Min.2d.inst.

. Notched
e it

y

6//I Snipped

**■ s/n Ragged
m 4/u Antlered

vm

3/n Strap
EC

?/!/ Vestig/af

X* '/it No wing

3-8 £3

s^

s S

■54 »t f

«■ g

Fig. 20. — Graphic representation of the action of the different ^-alleles upon
destruction of wing area. Abscissa instars; ordinate wing growth. (From
Goldschmidt, 1937, Univ. of Calif. Publ. Zool. 41.)

insure normal differentiation. In the Wild type and in all
compounds producing the Wild type, the rate of production of this
substance has to be above this minimum. As the production
must keep pace with the normal growth of the wing Anlage, the
top curve in the diagram might represent as well the growth
curve of the Anlage as the curve of production of the growth
substance. As the facts of development show, the deficiency of
this substance begins earlier and earlier in the progressive series
of alleles and continues to increase almost up to the end of wing-
development, as shown by the continuous degeneration. The
curves for the production of this substance must therefore begin
to drop at a certain point in development and continue to do so

60 PHYSIOLOGICAL GENETIC8

to the end. If the size of the Anlage at each moment of normal
development is given by the normal growth curve on top, the
ordinates of the other curves ought to indicate also the size of the
degenerating Anlage at different times, in fractions of the normal
size at the same time. The curves for the different alleles have
really been drawn according to the values of these fractions,
actually found and noted in the diagram. The resulting pheno-
type is marked on the right side. The diagram, then, represents
all the facts mentioned in terms of rates of production of some-
thing and apparently this general form, which might be varied
considerably in details, is the only consistent method of repre-
senting all the facts. (According to Kirchhoff's classic definition,
an explanation is the shortest description of all the facts.) As
mentioned before, the same facts may just as well be presented in
terms of production of something that destroys wing tissue. If
we accept this reciprocal type of representation, the curve (Fig.
20) means, besides the growth curve, the curve of production
of the substance in question which from a certain moment on
destroys tissue. This representation seems at first sight less
probable. But a most remarkable parallel is available, which
makes it worth while to investigate whether there is not some-
thing more than a superficial resemblance behind the two
phenomena. In a series of papers on the kinetics of bacteri-
ophage, Krueger (see 1936) found the following facts:

If a mixture of phage and growing bacteria is kept, the increase
of phage depends upon the bacterial growth in so far as it does
not occur without the latter. But the rate of increase of phage
is considerably greater than the rate of bacterial reproduction,
and therefore the ratio of phage to bacteria is constantly increas-
ing. When a certain ratio in favor of phage is reached — the
lytic threshold — lytic destruction of bacteria begins and proceeds
rapidly to completion. The time of onset of this destruction is
proportional to the initial concentration of phage in the mixture.
Phage may be inactivated and reactivated by differentiations.
In addition, in the presence of manganese, the lytic action begins
earlier; this is solely an effect upon the threshold, i.e., the quantity
of phage /bacterium requisite for lysis. Figure 21 represents
Krueger's graph for these experiments, which shows a perfect
model of the vestigial case. The growth curve of bacteria
corresponds to the normal wing growth in Drosophila. The same

THE MUTATED GENE

61

curve might also represent the increase in phage if it would be
proportional to the bacterial growth, as has been silently
assumed in the vestigial case for the production of the decisive
stuff. This assumption, which works in case of a growth sub-
stance, would not work in the case of production of a lytic sub-
stance, as we must account for the different time of onset of

0 I 2 3 4 5

Hours
Fig. 21. — Graphic representation of bacterial'growth, phage production and
lytic ratios with and without MnCl2. Originating in the bacterial growth curve o

are the curves of bacterial lysis with MnCh ( — ) and without ( ).

Above are two lines paralleling the logarithmic phase of the growth curve. Both
are crossed by a series of steep curves representing phage formation for various
initial phage concentrations. The intercepts of these latter lines with the two
parallel lines indicate attainment of the lytic phage bacteria ratios requisite for
lysis. At corresponding time intervals on the curve for bacterial growth the
curves of lytic destruction of bacteria begin. {From Kroeger and West, 1935,
J. Gen. Phys. 19.)

action. In Krueger's diagram, the different times of onset of
lytic action and the consequent dropping of the curves for bac-
terial growth are visible, and their similarity to the curves for
wing growth is obvious. Here we find, in addition, the steep
curves for growth of the phage drawn for four different initial
concentrations. Their intercept with a line parallel to the growth
curve, representing the assumed ratio of quantity phage/
bacterium necessary for lytic action, i- then the lytic threshold.
In the vestigial curve, this threshold has been indicated directly,
but it might as well be represented as in Krueger's graph. The

02 PHYSIOLOGICAL GENETICS

Upper parallel indicates the lytic threshold in the presence of Mn.
A parallel to this would be the action of the dominigenes (or
temperatures) upon the up-curves.

It is clear that we could use the graph for the phage also to
describe the vestigial case, viz., under the former assumption that
the genes produce here a lytic substance which acts upon the wing
as soon as a threshold value is reached. The main curve of
Krueger's graph would then represent the growrth and subsequent
lysis of the wing; his curves for phage production wrould be the
curves for production of the lytic substance; the initial concen-
tration of phage would be the different alleles of vg, and the
manganese curves would represent the action of modifiers which
will be studied later. The phage model then contains exactly
the system that we postulated for gene action and especially the
vestigial case: a growth curve; an onset of destruction at definite
times ; a proportion between this time and the basic concentration
(see later the theory of multiple alleles); and threshold values,
which might be modified by other substances produced by modi-
fying genes. The parallelism, if not more significant, shows at
least one thing: The type of genie actions as derived from the
author's work is actually found in comparable processes in nature.

It ought to be added that Mohr in his study of this allelo-
morphic series drew some similar conclusions, in so far as it was
possible without knowledge of the development. He speaks
of the different potencies of the alleles in producing the respective
amounts of scalloping and assigns definite numerical values to
these potencies, a method of description that was introduced by
Goldschmidt (1912) in his work on the potencies of sex genes.
He gives the highest value to the potency for production of the
Normal wing (30 for one gene) and the lowest (0 to one No-wring
gene) as estimated, of course, from the phenotypical result.
These values are found at the base of Fig. 19. The threshold
for the first indication of a Nick, then, is found at potency 37.
He discusses further the probability that these potencies act by
controlling reaction velocities.

It is probable that similar cases may be found in plants, as
might be inferred from Van Overbeek's (1935) work. He
studied the recessive character nana in corn which causes an
inhibition of growth of the mesocotyl. He studied the behavior
of the growth hormone (auxin) in this mutation and found that

THE MUTATED GENE 63

nana gives off less auxin from the tip of the coleptile and also
that it grows less with the same amount of auxin. It was found
that the reason for both is that in nana auxin is destroyed at a
higher rate. The reason for this seems to be a change in oxida-
tion reduction. In this case, the result is only dwarfism, but a
similar effect might in other cases lead to tissue destruction, as
in the vestigial wing.

There is one case existing in plants that falls in line with the
facts discussed in this and the foregoing chapter and has been
used also to derive a corresponding explanation. Oehlkers
(1930-1935) analyzed the Cruciata character of Oenothera.
This hereditary pathological trait might be compared to cases of
Homoeosis in Drosophila, as homologous organs replace each
other. Cruciata flowers are sepaloid, the petals being more or
less transformed into sepals. All transitions exist from petals
with a trace of sepaloidy through different mixtures of petal and
sepal tissue to pure sepals. The trait (which had already been
studied by de Vries and others) behaves as a simple Mendelian
recessive. The grade of sepaloidy is also inherited, is typical
for each Oenothera complex, and remains constant in crosses.
The phenotypic effect of different combinations follows simple
quantitative rules and is such as expected, if each of these
Cruciata genes — assuming them to be members of an allelo-
morphic series — has a definite quantitative effect in regard to the
Cruciata character. The case therefore resembles most closely
the case of the sex genes in Lymantria, where also a series of
different conditions of the gene exist, which in all combinations
have the proper additive effect which can be calculated from the
individual effects. Oehlkers represents his facts, then, using the
type of the diagram introduced for Lymantria and in terms of
assumed quantities of action of the respective genes. His
diagrammatic representation of the facts takes as a basis a
quantity of 10 for one definite complex with complete sepaloidy
and of 42 for one with complete normality. Parallel to the case
of intersexuality in Lymantria, it is assumed that everything
above 26 is normal, below 16 sepaloid, and between 16 and 26
intermediate in five classes (Fig. 22). The different combinations
of complexes with Cruciata must then give consistent results if
definite quantities are assigned to the individual actions. This
is the case.

64

PHYSIOLOGICAL GENETICS

It has to be added, however, thai in this case the presence of a
series of multiple allele- is made probable bu1 qoI s1 rictly proved,
as the special conditions of complex heterozygosis in Oenotheras

-- 46
-- 45
-- 44
-- 43

42 Cn ■ Cn AHookeri • ''Hookeri from I ><•. Bookeri.

4!

40 Crj ■ Crs albicans • flavens from ()e. suaveolens sulfurea.

39

38

37

36

35

34

33

32

31

30 Cr3 • Cn[albicans • albicans from Oe. hie. sulfurea Hannover].

29
-- 28

Kl.W__

-- 11

26 5 Cn • Cn [rubens • rubens from Oe. bi. sulfurea Hannover].

26'

2S.5

25

— 24 cn • cn [albicans ■ albicans from Oe. bi. cruciata A 2a].

jzt + 23

22 cn • cn Irubens • rubens from Oe. bie. cruciata A 2a].

1¥ 4 21

20.5 cn • en [gaudens • gaudens from Oe. Lam. cruciata].

19
18
17
16

15
14
13
"" ]?, cn • cn [velans velans from Oe. Lam. cruciata]

10
9

en ■ crt albicans ■ rubens, Oe. bie. cruciata apetala.

y

Fig. 22. — Graphic representation of the relative effects of the genes Cr — cr.
In brackets, combinations that may not be obtained. Kl I -VII, phenotypic
classes; cr 1-5, cruciate genes of increasing potency. [From Oehlkers, 1930,
Ztschr. f. Bot. {Oltmann Festschr).}

make a complete proof rather difficult. It would be interesting
to try to formulate a more detailed view of genie action in this
case. Unfortunately, the basic facts concerning developmental

THE MUTATED GENE 65

mechanics of the case are unknown. A zoologist would suspect
that the normal differentiation of sepals and petals has to do
with a process that we shall later discuss as stratification of
determining substances (which occurs a long time before visible
differentiation). Sepalody would then be the effect of insufficient
segregation of such stuffs in the flower bud. If this were the case,
an interpretation parallel to the wjr-case involving velocities of
definite processes might be developed easily. Unfortunately, the
physiological facts are lacking. But there is one significant fact
strongly pointing in the direction of such a general explanation:
Oehlkers (1935) found that the Cruciata gene increases the rate
of differentiation of the plant and made it even probable that this
increase is proportional to the relative strength of these
allelomorphs.

2. The Threshold Concept. — Before reporting further cases
of the same type, we must consider the threshold concept, which
we are facing here for the first time and which will reappear on
many occasions. The diagram (Fig. 20) shows in the curve for
the Wild-type wing (on top) that the production of the decisive
substance always has to be above the necessary minimum, the
threshold, to insure normal development (or, in the reciprocal
conception, below the threshold). We could imagine that many
more such curves above the normal curve exist, all of which
would be above the threshold and therefore control different
degrees of Wild-type, viz., plus minus hyperwild. We shall return
to this special point in the chapter on dominance and shall discuss
here only the general features of the threshold concept. Early
in the development of the theory of genie action by rates of
reaction, it became clear that the explanation of many details
requires the assumption that the products of those reactions
which might generally be called determining stuffs act only when
they have reached a minimum concentration, i.e., a threshold
level. To this may be added the conception that there is also a
maximum of response: Above a certain level of concentration, no
increase in response is possible; e.g., an increase in the wing-
growth substances will not change the effect, the maximum of
which is the Wild-type wing. This level of possible maximum
response may coincide with the minimum threshold, as in the
case just mentioned, or it might also be higher. We shall meet
later with examples.

66 PHYSIOLOGICAL GENETICS

The threshold conception had already been used by Gold-
schmidt in his firsl analysis of the phenocopic effect (19206), as
well as before, in his first analysis of intersexuality (the epistatie
minimum, Goldschmidt, 1912, 1920c). Wright also (1916) had
realized its importance in explaining the action of the pigmenta-
tion genes in guinea pigs (see page 88), which he conceived in a
way similar to that presented in Goldschmidt's simultaneous
work. It has since been used considerably by many authors who
discuss questions of rate (Crew, Haldane, Huxley, Plunkett,
Wright, Mohr, et al.; see also its interesting application to certain
problems of sex determination by Dobzhansky and Goldschmidt,
1934c). Wre shall meet it over again in connection with different
problems. Here only one of its aspects will be mentioned, which,
though highly important, has not yet been analyzed sufficiently.
In Mohr's table (page 57), we found that the lowest members
of the v ^-series, which produce only a small nick in the wing, do
it not in all individuals but only in a small percentage. In
higher gene combinations, this percentage increases finally up to
100 per cent, parallel with a quantitative increase of scalloping.
Here is a very important problem from the standpoint of gene
action.

Timofeeff-Ressowsky (1926) has called this peculiarity of the
action of some genes or alleles, which in Drosophila melanogaster
is typical for some lower alleles in a series and which is in other
species of Drosophila the most frequent form of gene expression.
the penetrance; this term might sometimes be useful for a short
description of such cases. An explanation of the phenomenon in
terms of threshold is suggested, when this phenomenon appears
as a member of a simple quantitative series of effects at one end
of the series, as in the vg-ca.se. Here a further feature is also
added, viz., a parallel increase of the percentage phenotypic
effect (penetrance) with the increase of the degree of the phe-
nomenon, in this instance the amount of scalloping. (Timofeeff
calls this the expressivity, a term that might lead to rather
dangerous misunderstandings.) Such a series of facts of an
orderly linear nature requires an explanation of the same type.
This is easily conceived if the threshold concept is applied. To
show that t he same general type of interpretation may work with
different conceptions in detail, we may apply two such concep-
tions to the vestigial case as follows:

THE MUTATED GENE 67

1. The time at which an event leading to degeneration is still
possible is limited, as we saw, to a definite stage in wing differ-
entiation. This time would then constitute a limiting value.
The action of the gene in question may require so much time to
reach the necessary concentration for action, or, in the reciprocal
conception, the curve of production of the growth substance may
drop below the threshold at so late a time in development, that
in either case the event approaches the limiting time. The normal
variation in the time of developmental processes may then cause
a larger or smaller part of the curve of variation of developmental
time to cover the time at which the gene-controlled process
takes place. The result is that all individuals belonging to the
part of the curve beyond the limiting value for possible changes
remain normal.

2. The same result would be obtained if the time of differentia-
tion remained constant and the moment at which the curve drops
were variable, or if the threshold concentration for the stuff in
question were variable or another member of the system or some
or all would vary.

In general terms, then, the phenomenon would result from the
general fact of fluctuation of the particular processes of develop-
ment in conjunction with the existence of thresholds for the
integrating processes.

An actual example of this type in which some of the elements
are known is the following: In Lymantria dispar, male intersexes
of different grades may be produced by combining in appropriate
crosses sex genes of typical but different valencies. If, for
example, a constant female determiner of high valency, a so-called
strong F (which has turned out to be a cytoplasmic quality), is
combined with one constant male determiner of low valency or a
weak M (which is inherited within the X-chromosome), the second
M in the male formula (2X = <? ) may be taken from races of
different valencies. In the balanced system of sex determiners,
which might be written F /MM, which is the general formula for
males, the valencies have been combined as Fstr/MwMvaT, Mvar
meaning that this M alone will be varied in different combinations
built up by crossing. The respective valencies of M in different
races are known from other experiments. Introducing then
M-genes of the highest valency (= strong) through intermediate
stages to the lowest valency (= weak) into individuals of this

68

PHYSIOLOGICAL GENETICS

formula, the sexual balance is shifted correspondingly in favor of
F. The result is thai an ascending scries of male intersexes ie
produced, beginning w itb a combination giving normal male- and
Leading through all Btages of low to high intersexuality, finally to
sex reversal, i.e., females. The actual result of such a series of
replacement experiments is represented in the following table.

Table

5.

Strong F
from race

Very weak M
from race

Second M of
increasing
strength
from race

0?

Intersexual o* grade
(black line = range of variation)

? (genetic </)

I

II

III

IV

V

VI

Tokyo

Hokkaido

Hokkaido

—

Tokyo

Hokkaido

Berlin

Tokyo

Hokkaido

Russia

Tokyo

Hokkaido

Korea

Tokyo

Hokkaido

Kumamoto

Tokyo

Hokkaido

Kyoto

Tokyo

Hokkaido

Tokyo

—

Here we notice the phenomenon that is of interest in the present
discussion. If the sexual balance is shifted so that the lowest
grade of intersexuality is produced, it does not happen to all
individuals but only to a certain percentage, which increases with
increasing grade of intersexuality. There is obviously a thres-
hold, or limiting value, for F/MM at which intersexuality
begins (called the epistatic minimum in this case). And there is
obviously a fluctuation of the value F/MM around a mean —
whether as a modification or produced by modifying genes is
irrelevant — with the consequence that only individuals belonging
to the part of the range of variation beyond the threshold become
intersexual. (Pictures of such a series of male intersexes are
given in Fig. 35, page 194.) The same will be found at the other
end of the series, where a high degree of intersexuality fluctuates
into femaleness. It might be added that recent discussions of
the special problems of sex and intersexuality by Dobzhansky and
Goldschmidt (see page 66) have added the possibility that the
cytoplasmic factor which is designated as F, because working
in the direction of female determination, might in fact be some-

THE MUTATED GENE 69

thing that determines the position of the threshold for the suc-
cessful action of the M-genes within the X-chromosomes. It
may be added, also, that the other example mentioned in the last
chapter, Oehlker's Oenotheras, show the same threshold
phenomenon (see Fig. 22). At the threshold limits (26 for
normal versus low-grade sepaloid), a fluctuation of some indi-
viduals across the border line occurs. Furthermore, a number of
genotypes are found of different quantitative value, pheno-
typically alike because located beyond the threshold, and strictly
comparable to the hypermales and females in Lymantria and the
hyperwild in vestigial flies.

3. Penetrance. — The facts reported on pages 65 to 68 brought
us again into contact with a problem that had been occasionally
mentioned before, the discussion of which finds its proper place
here. Timofeeff (1926) introduced the term penetrance to
describe the fact that the phenotypic effect of a gene does not
always take place in 100 per cent of the individuals homozygous
for the gene. In the vestigial case (see Fig. 14), the genes vgni/vgni
had zero penetrance; and the nicked-notched compound,
0.2 per cent penetrance. This term might be used for descriptive
purposes, provided one realizes that penetrance is a phenotypic
result, which may have many underlying reasons such as the
absence of genetic modifiers necessary for the production of the
effect, the existence of a threshold for visible effect which is con-
trolled genetically or by outside conditions, or a weak potency
of the gene in question, combined with a threshold that is not
reached. Of these possibilities the first one may be omitted
here, for if Mendelian segregation of modifiers, present in the line,
leads .to definite percentages of visible effects, it amounts to a
question of Mendelian segregation with genes of otherwise
invisible effect. If such modifying genes, however, control
threshold conditions, the problem is the same as all threshold
problems in regard to the details of explanation. But the numer-
ical values of penetrance are again the result of a Mendelian
segregation, this time of modifiers for the threshold of visible
action of the main gene.

In the preceding section, we mentioned the cases in which
obviously a low percentage of penetrance was the result of the
presence of a threshold in such a position that the always present
fluctuation of the decisive process allowed only a small number

70 PHYSIOLOGICAL GENETICS

of the individuals t<> reach the threshold. In those cases, the
position of the threshold was determined by the quantity of a
determining substance or the time of its onset. It may be safely
assumed that in most eases of penetrance this is the actual
situation. But it must be kept in mind that in each case the
action of modifying genes will first have to be excluded, before
any conclusion upon gene action may be drawn from the material.
We may conveniently group the material into the following
types: (1) genes that under normal conditions do not manifest
their action but have a visible effect under other definite condi-
tions; (2) genes that under standard conditions (and excluding
modifiers) produce a visible effect in a certain percentage of the
individuals containing the gene. It is understood that in such
cases the normal individuals must be proved to have the same
genetic constitution as the others. In this case, genetic or
environmental conditions may be shown to control the percentage
of changed phenotypes. Examples for these types of behavior
are abundant in genetic literature. We shall report only those in
which experimentation has yielded information relevant to the
problems under discussion.

There is the frequently quoted case of reduplication of legs in
Drosophila due to a sex-linked mutant gene. Hoge (1915)
found that this type appears only in a certain percentage of the
homozygous flies but that if reared at a low temperature the per-
centage increases from between 10 and 15 to between 30 and 68
per cent. The action of low temperature was found to be limited
to a sensitive period which is situated very early in development.
This is, of course, to be expected, as the action must take place
before the imaginal disks of the legs are finally determined. One
might conclude from the facts that the action of the mutant gene
takes place at a time very near to the time of final determination
of the Anlage, a time limit beyond which a fluctuation occurs
depending upon temperature (see end of this section, page 72).
Similar results have been reported by Morgan (19156) for the char-
acter abnormal abdomen where humidity is the decisive agent, by
Komai (1926) for crippled legs in Drosophila, by Astauroff (1930)
for the mutant tetraptera. (Here the penetrance increases from
1 to 35 per cent with a rise in temperature from 17 to 27°C.)

Of a somewhat different type is the work of Wright (1934a) on
extra toes in guinea pigs. This character is not the result of a

THE MUTATED GENE 71

single mutation but seems to be caused by a major differential
factor and some additional minor factors. By selection and
inbreeding, different lines were isolated, characterized by different
percentages of penetrance of the character. In this case, it was
also proved that the normal individuals had the same genetic
constitution as the four-toed ones. Some of the facts pointed
in the direction of external influences upon the percentage inci-
dence of the type, but mammals are unsatisfactory material for
such experimentation. Wright realized also, just as Goldschmidt
had in the cases reported before, that the problem comes down
to the question of a physiological threshold, which is reached or
not reached under certain combinations of genetic as well as
environmental conditions.

A systematic inquiry into this problem has been made by
Timofeeff (1925, 1929a, 6, 1934) who worked with Drosophila
funcbris, a species in which most of the mutants do not show a
100 per cent effect but a lower percentage of penetrance under
standard conditions. The mutant in question is venae transversae
incompletae (vti) of D. funebris, exhibiting an interruption or
absence of the cross veins of the wing. Vti is an autosomal reces-
sive which requires in addition the presence of another gene for
visible manifestation. Furthermore, there are present other
modifying genes which influence the percentage of individuals
exhibiting the defect, i.e., the penetrance. By means of selection,
lines could be isolated which showed from 41 to 100 per cent
penetrance, respectively. (There could also be isolated lines
differing in the quantitative amount of the defect and in its
specific pattern and symmetry. These facts do not pertain to the
problems of this chapter.) With these selected lines, experiments
were performed which proved an influence of temperature action
upon the percentage manifestation, but only in some of the lines.
In general, the percentage of manifestation decreases with increase
in temperature. Thus 13°C. gives 97 per cent, and 28° gives
74 per cent if the temperature is applied during a sensitive
period in the first half of larval life. But there is another sensi-
tive period in pupal life where temperature acts reciprocally; a
rise in temperature produces a small increase in penetrance.
The explanation that Timofeeff gives for these facts falls in line
with all the general conclusions that we have presented in former
chapters. During the early sensitive period, the processes that

72 PHYSIOLOGICAL CKNETICS

are necessary for the cross-vein format ion, e.g., the formation of
morphogenetic Bubstances, are speeded up by temperature action.
In the second sensitive period, coinciding with the actual forma-
tion of the veins, the amount of these substances is determined,
but the time of development may be differentially changed by
temperature action. This is, of course, the same type of interpre-
tation thai has always been found for gene-controlled processes,
following Goldschmidt's theory of timed velocities.

But the decisive point has not been touched here: why is the
effect produced in different percentages of individuals? This
part of the problem requires, no doubt, some form of the threshold
concept combined with the conception of a certain fluctuation
of the effect, as derived above for the cases studied by Gold-
schmidt. If this fluctuation takes place near the level of the
threshold, a more or less considerable part of the range of fluctua-
tion may fall above the threshold. We have already emphasized
this point in the last section. This shift of the curve of fluctua-
tion of some decisive process beyond the threshold may, however,
be brought about by an action of temperature or of so-called
modifying genes, e.g., as a shift of the mean of fluctuation toward
or away from the threshold value; or as an increase or decrease
of the range of fluctuation of this reaction product; or as a shift
of the threshold value for the action of that substance. Then
penetrance, as far as it is not hidden segregation, is a phenomenon
of threshold combined with statistic variability of the products
of gene action. In a recent paper on the scute gene, Child
(1936) applies the same concept to the process of bristle formation
and illustrates it by the same type of diagram that Goldschmidt
used in other cases.

4. Further Evidence. — We begin with work that was under-
taken to find out whether an insight into the action of the mutant
gene could be obtained by the use of quantitative methods.
Some of this work has been reported in earlier chapters (see
pages 31-34), or parts of it have been mentioned in other con-
nections.

The first author who had derived the rate concept from experi-
ments on somatic characters produced by a series of multiple
allelomorphs was Wright (1916). Almost simultaneously Gold-
schmidt derived a rate concept from his work on the multiple
allelomorphs of sex genes (19176) and pigmentation (1917d),

THE MUTATED GENE 73

which led to a generalization of the idea. Wright found a
serievS of allelomorphs of the albino type in guinea pigs, which
produced an orderly effect upon skin color and which could be
seriated. The general nature of the effects suggested that the
albino series has to do simply with the rate of a process that is
essential to all melanin production. But a difficulty was met
when the effects of the series of genes were studied in relation
to the underlying constitution of black fur, because it was found
that the order of effects was not the same. But this might have
been due to secondary processes in the physiology of pigment
production, e.g., threshold conditions. We shall return to this
special point in a later chapter (see page 88), where the details
of the case will be reported. Here it suffices to say that a later,
more complete analysis (Wright, 1925) led the author to confirm
his earlier views that the factors of the albino series determine the
rate of some one process fundamental to all pigmentation.

In Drosophila, Plunkett (1926) first applied the concept of
gene-controlled rates to the results of his experiments. The
school of Zeleny (see page 75) had already shown that a gene-
controlled character, the number of facets in the Bar eye, varies
reciprocally as the temperature at which the flies were raised.
They expressed the results in terms of the temperature coefficient
for the rate of facet formation but did not conclude that the
mutant gene itself acts by changing the rate of some process, as
Wright and Goldschmidt had done. Plunkett used for this
experiment the gene Dichaete which removes certain bristles on
the thorax of Drosophila, especially the dorsocentrals. This
effect, as usual, is influenced by modifiers and external conditions,
and it reacts most regularly to temperature changes and, accord-
ing to Plunkett, without a sensitive period. Table 6 gives the
data for mean numbers of the posterior dorsocentral bristles on
one side of the body (1 in the Wild type).

Plunkett analyzed these data mathematically. As only a
single bristle is involved which is produced at a definite time, the
concentration of a bristle-forming or a bristle-suppressing sub-
stance must be involved.1 The effect may then be produced
either by an increase, relative to other developmental processes,
in the rate of a reaction, breaking down a bristle-forming sub-

1 It is not known whether these mutants prevent bristle formation or
destroy already formed bristles, as is the case in vestigial wings.

74 PHYSIOLOGICAL GENETICS

Table

6

From Plunkett)

Degrees

Centigrade

M

14

0.690

15

0.600

17

0.581

20

0.439

21

0.364

24.5

0 . 266

25

0.252

27

0.124

30

0.054

stance or forming a bristle-suppressing substance; or by a decrease
relative to the other developmental processes in the rate of such
reactions. (Note the parallel to the vestigial case). It is known,
of course, that the general rate of development increases with
increase of temperature, and therefore the bristle-forming reac-
tion must have a specific temperature coefficient (as had been
assumed by Krafka, 1920a; Goldschmidt, 1920b; and others in
their cases). The fact of increase of bristle-destroying effect
with temperature points to the second alternative — decrease
of rate relative to general development. It is then assumed that
the concentration of the bristle substance is proportional to
the mean number of bristles produced and that the rate of the
decisive reaction that reduces the bristles also is always propor-
tional to that concentration. The total time of this reaction is
calculated to be proportional to the total time of development at
all temperatures within the vital optimum. Now, the difference
between Wild type and Dichaete is that in the latter a reaction
takes place that may be described as bristle destroying. This
reaction of the Dichaete gene, then, is to produce a catalyst which
catalyzes the bristle-destroying (or obstructing) substance with
a definite velocity proportional to its concentration. (Note,
again, the parallel to the vestigial case and the phage.) From
these premises, Plunkett proceeds to calculate the thermal
increment for the bristle-reducing reaction on the basis of an
irreversible monomolecular reaction, using the data for tempera-
ture-bristle effect and those for time of development. Further-
more, he calculates the effect of different temperatures as expected
if the equation that was used actually applies. The calculated
expectations fit very closely the data of the foregoing table. The

THE MUTATED GENE 75

details of these calculations, e.g., value of n, point to the assump-
tion that the reaction in question, produced by the Dichaete
gene, is the decomposition of a thermolabile catalyst. And
whereas the effect is bristle or no bristle, there must also be
involved a threshold value for bristle- removing power. The
Dichaete gene, then, is assumed to be a catalyst which catalyzes
a chain of reactions which accelerates the (in itself spontaneous
but otherwise infinitesimally slow) decomposition of a thermola-
bile bristle-forming catalyst by catalyzing the production of a
destructive catalyst. It must finally be assumed that all genes
act essentially in this way. Although he comes to just the same
conclusions as the aforementioned authors, in this analysis
Plunkett has delved further into the details of chemical kinetics
than any other investigator and has shown that the rate concept
of genie action also stands the test of subtle mathematical treat-
ment. It might be added that the kinetic side of this work agrees
again most remarkably with the work on the bacteriophage, but
the actual embryological processes are not known as in the
vestigial case.

Considerable work has been done with the eye mutants of
Drosophila, which lend themselves to quantitative treatment
because the number of ommatidia is a suitable numerical char-
acter. We have discussed Zeleny's and Krafka's work, which
has been carried further by Driver and Hersh, whose experi-
ments on temperature action and the sensitive period were
reported. In these experiments, the following additional results
were found (Driver, 1931). The temperature effect, viz., the
increase in facet number per degree of decrease of tempera-
ture, follows a geometrical formula which Driver expresses as
(1 -f- rn) = Fi/F2, where r equals the rate of change, n the number
of degrees decrease in temperature, Fi the facet count at the lower,
F2 that at the higher temperature. This progressive effect
shows a definite change in rate at 21° in Bar flies but not in
Ultrabar. This action of temperature occurs during the sensitive
period; but the length of this period (as percentage of larval life)
in this case is not the same at all temperatures, becoming smaller
with increase in temperature. Ultrabar, however, does not show
this variation but shows a change with the temperature, of the
onset of the period relative to developmental stage. In addition,
it must be stated that the normal eye shows very little of all

76 PHYSIOLOGICAL GENETICS

these temperature effects. An analysis of these facts was made
by trying to fit them to the different formulas for temperature
action upon chemical processes. They did not fit either of
them, and Driver therefore concludes that the series of Bar
genes affect different simultaneous reactions differentially —
either retarding the rate of facet formation or changing the limits
of the effective periods or both more or less — and that these
changes, in addition, do not parallel the simultaneous changes of
rates of general developmental processes (differences in tem-
perature coefficient). Within the sensitive period, the rate of
facet formation seems to progress evenly. Altogether, then,
the facts point to an action of the mutant gene upon the rates
of closely interwoven but simultaneous reactions, as postulated
in Goldschmidt's theory of development. A better knowledge
of the development of the Bar eye might help to a better under-
standing of the processes involved. Chen's (1929) findings that
the very young Anlage of the Bar eye is not smaller than a
Wild-type eye points to a process parallel to the one reported
for vestigial. We shall meet the same set of facts many times
in later chapters, where additional facts will be reported.

But one more recent investigation is to be mentioned because
it will possibly reduce the Bar case to the same type of processes
that we described for vestigial. In a paper dealing with the
action of the t>o-gene upon the expression of the Bar series,
Margolis (1935a) finds evidence that actually the effect of the
5-gene may have nothing to do with the process of formation
of ommatidia but is concerned rather with the production of
something that destroys the material for ommatidia (see also
Wolsky, page 32), which would bring the Bar case into line with
the vestigial case. He found that the presence of vg, which,
as we know, retards development, decreases the number of
facets. This would be the case if a facet-destroying action of
Bar were prolonged by lengthening the time of its action. All
the data presented above would be interpreted in terms of a sys-
tem of reactions in time, which Margolis develops tentatively
along lines similar to those developed by Goldschmidt in his
analytical work.

In a quite recent paper (Margolis and Robertson, 1937),
Margolis modifies his interpretation somewhat. Certain irregu-
larities in the behavior of the sensitive period in Wild flies are

THE MUTATED GENE 77

interpreted as due to a collaboration of facet-forming and facet-
destroying processes. The catalytic action of the Bar gene might
then mean an acceleration of the destroying processes. This
interpretation, of course, does not change the principle involved.

It is not possible to record here all cases in which the applica-
tion of the rate concept to gene-controlled processes has led to a
better understanding. We shall have to discuss some further
examples in connection with other problems. Only a few more
cases may be mentioned as types of applicability of the concept
to various processes different in detail.

A typical gene-controlled character with a visible action in
time is the molting of insects between the different instars. It
is known for the silkworm (Tanaka, 1916; Ogura, 1932) and for
the gypsy moth (Goldschmidt, 19246, 1933a) that races exist with
different numbers of molts (3 to 5 in the silk worm; 4 to 5 in
Lymantria). The genetic difference involves a series of multiple
allelomorphs controlling the different types. Their effect may
be shifted by the presence of modifying genes (Ogura) and also
by external conditions like feeding with coal dust (Nagai)1
or hunger (Nagamori; Goldschmidt).2 From Goldschmidt's
work on the growth curves of such races, it is clear that the num-
ber of molts is determined independently of growth in general.
The last molt, pupation, of course sets an end to further molts,
and this is determined by a definite concentration of a hormone
produced in the corpora allata near the brain (Wigglesworth).
Ogura, who made a detailed genetical analysis of the case in
the silkworm, came to the conclusion that all his facts could
be harmonized by the assumption that the different allelomorphs
control the rate of production of this hormone, which acts when a
threshold value is reached. Modifiers and external conditions
will then act in the same way as described above for other cases.

Of a very different type is the following example which shows
how the rate concept accounts for complicated morphological
changes. In analyzing the corkscrew type of deer horns,
Rhumbler (1929) found that the normal growth requires the
exact timing of two processes, viz., the rate of growth of soft
preosseal connective tissue and the rate of the ossification ascend-
ing from below. If the latter process has too low a rate, the

1 Quoted from Ogura.

2 Unpublished.

78 PHYSIOLOGICAL GENETICS

corkscrew horn will result by the action of gravity upon the not
sufficiently supported tissue

Only one more example might be added from a different field
of knowledge to show how many problems might be approached
and coordinated with the rate concept. I mention the work of
Ki'ihne on the hereditary abnormalities of the human vertebral
column and its interpretation by Eugen Fischer (1933). It
was shown that such abnormalities are inherited in a simple
Mendelian way. It is not the individual variation that is
inherited but a general tendency to shift the limits of the sub-
regions of the column anteriorly or posteriorly. In identical
twins, the type of shift is identical. A phylogenetic shift, com-
parable to these, may also be observed when comparing different
types of apes and monkeys. Fischer has tried to explain these
facts with a rate concept by assuming that the rate of differentia-
tion of the vertebrae is decisive. For example, in all men a
thirteenth rib is primarily formed, and the twenty-fifth vertebra
is still laid down as a lumbar vertebra. A shift of the rate of
differentiation from tail to head makes the Anlage of the thir-
teenth rib concresce with the twentieth vertebra and disappear;
a still higher rate may similarly affect the twelfth rib. Corre-
spondingly, the tw^enty-fifth vertebra is incorporated into the
sacrum; and vice versa, a slower rate of differentiation (= con-
crescence) will leave the Anlagen free. In a general way, it is
thus possible to account for the facts without assuming special
localized actions. In addition, some phylogenetic facts may be
reduced to as simple a thing as rate differences (see also page 211).

4. SOME SPECIAL TYPES OF MUTANT GENES AND GENIC ACTIONS

The elementary facts that have been reported thus far are
being supplemented by a considerable body of evidence which
has been obtained in the study of special cases of behavior of
genes and of their action. This evidence will be presented in
the present chapter.

A. Pleiotropy

This term is frequently used to describe the fact that a given
mutant gene produces not only a single change, the one that is
usually attributed to it, e.g., vestigial wing, white eyes, but also
a number of different phenotypic changes, the number of which

THE MUTATED GENE 79

usually increases with more intimate knowledge. We may
consider a few examples. The vestigial gene in Drosophila
produces as its main effect the typical reduced wing. But it
also causes the scutellar bristles to point forward and upward,
prolongs the time of development, decreases viability, makes
wings divergent, and causes rudimentation of balancers. The
gene for red eyes in the flour moth simultaneously makes the
testes colorless, the skin of caterpillars pale, the optic ganglia red
instead of brown, the ocelli of the larvae less pigmented, and
decreases the speed of development and the vitality. Numerous
examples of this type may be given, some of which we shall
discuss in relation to special problems. These effects are usually
described as a consequence of the fact that any character is
controlled by all of the genes and that only the relative effect of
one gene may be larger in a given case, thus making it possible to
attribute one major effect to one gene. Such a statement may
be correct, but it is not necessarily so. This depends completely
upon the developmental process which is changed by a mutant
gene and upon the time of incidence of this process. If, for
example, a mutant gene like vgni produces a deficiency of some
growth substance toward the end of development of the Droso-
phila wing, no other effects may occur, or only such simultaneous
or later differentiations may be affected as are dependent upon
the quantity of the same substance (if any). If, however,
the allele vgNw produces a similar deficiency at an early stage of
differentiation of the wing disk, it is very likely that other
processes of growth that are intimately correlated with imaginal
disk differentiation at that stage will be influenced, too, resulting
in slower development, diminished viability, and eventual
morphological consequences. One of the latter can actually be
shown to be a mechanical consequence of the disturbed growth,
viz., the upright position of the scutellar bristles. It frequently
occurs as a nonheritable abnormality that the two halves of the
notum (which are partly developed within the same disk as the
wings) do not concresce properly. In such cases, the scutellar
bristles may assume the same position as in vestigial, showing
that this position is a consequence of certain mechanical condi-
tions during concrescence. In vestigial flies, such disturbances
are more frequent than in Wild stock, and it may be regarded
as extremely probable that the bristle character is nothing but

so

/'// YSIOU )QICA I. GENETICS

;i mechanical consequence of the disturbance of the process of
concrescence, produced by the undersize of 1 he wing-forming part
of the imaginal disk. A still earlier onset of the pathological
process would, of course, have still more consequences; in this
case, they are not visible because such forms (homozygous
No-wing) cannot develop at all — are lethal.

The working of this interpretation may be exemplified by
considering Table 7. Mohr (1932) compared these manifold
effects of the vgr-gene in a series of allelomorphs and compounds.
(We must keep in mind that the increasing effect upon the wing
was proved to be the consequence of an earlier onset of the
destructive process.)

Table 7
(Extracted from Mohr)

Wing

Ab-
sent

Stumps

Ant-

lered

Rag-
ged

Scal-
loped

Notched

Nicked

Nor-
mal

Per cent incisions

Erect scutellars

Rudimentary balancers

100

+ +

+ +
+ + +

100

+ +
+ +
+ +

100

+
+
+

100

+ -
+ -

+ -

70.7

42.4

27-0.2

0

It should be added that time of development and viability
would give the same seriation of the effects, increasing with
earlier onset of the abnormality. We shall soon have to discuss
the same facts again.

There is also available some material in plants that leads to
the same conclusions. Anderson and de Winton (1935) found in
Primula sinensis that genes affecting the organs of the earlier
vegetative period (leaves) also affected flowers but not vice versa.
If by another mutant gene bracts became leafy, genes acting in
the vegetative phase also affected those bracts. Normal bracts,
however, react as bracts. Further examples of the same type of
action are given, though it is not clear how different types of
action upon different organs may be caused by a single genie
product.

This rule, that the pleiotropic effect is proportional, ceteris
paribus, to the earliness of onset of the gene-controlled change in
development, may also be illustrated from the cases of mon-

Til E MUTATED GENE 81

strosities in vertebrates which we studied before. But it must
be kept in mind that this time element is not the only cause of
pleiotropic gene effects. If the direct action of a gene consists
in the production of a substance acting like a hormone, this may
affect development all over the body. If the primary effect is
of a strictly localized type, nothing else may be affected, except
perhaps readjustments in the growth of neighboring parts.
Many intermediate and similar cases may be worked out theo-
retically or be derived from the study of the embryology of
mutants. The details will differ in the various cases. But it
will always be found that the so-called pleiotropic effect of a gene
is not a property of the gene and not a general property of gene
action but an embryological consequence of the time, place, and
type of the primary disturbance of development by the mutant
gene.

This explains also why the different characters controlled by a
mutant pleiotropic gene may show very different reactions.
An extreme case of this type has been reported by Eker (1935).
The recessive mutant short wing in Drosophila produces small
and rough eyes, short wings, scalloping, abnormal venation.
These characters have a strange relation to temperature: At 31°,
the viability is zero and increases with lower temperature. The
phenotypic expression of the traits is 100 per cent at 27.5° and
decreases to zero at 14°. (Of course, many cases are known of the
dependence of phenotypic expression upon -temperature or
moisture.) But in addition the different traits have a different
temperature-sensitive period — some during the whole of develop-
ment; others, only at a definite stage. (This case might, how-
ever, turn out not to be the result of a simple gene mutation.)

B. Multiple Alleles

The phenomenon of multiple allelomorphism, discovered by the
early Mendelian scholars, has always played an important role
in the study of gene action, since the Drosophila workers proved
that series of multiple allelomorphs are situated at the same locus
of the chromosome and therefore may be conceived as different
conditions of the same gene, whatever this may mean. As far
as I know, Wright (1916) and Goldschmidt (19166, 1917r/) were
the first to derive from the study of such a phenomenon general
ideas about the type of action of the gene. According to Wright,

82 PHYSIOLOGICAL GENETICS

t he action of the albino scries of alleles iii the guinea pig is to be
explained on the basis of tour quantitative gradations of one
factor, which determines the amount of the basic color-producing

enzyme. This is produced at a definite rate and acts only above
a certain threshold. Here, then, the conception of rate ami
threshold and of quantitative gradations is applied to explain the
details of a case of multiple alleles (see page 88).

Goldschmidt, as described on page 52, studied a series of
allelomorphs controlling pigmentation in caterpillars. Ho could
follow these effects through development and found that pigment
i- developed in different velocities in the different genotypes,
which could be characterized by their typical curve of pigmenta-
tion, From this it was concluded that multiple allelomorphs
act by controlling different rates of one and the same reaction.
The same view was simultaneously derived for the action of a
series of multiple allelomorphs of the sex genes. In this chapter,
we are not concerned with the conclusions upon the nature of
multiple allelomorphs but only with the problem of gene action.
We have already mentioned a number of cases in which the
effects of a series of multiple allelomorphs could be described in
such terms, e.g., the vestigial series. A priori, such a description
will be expected to apply only to those cases — the majority —
in which a series of multiple alleles produces an orderly series of
graded effects. Whether other types of multiple alleles exist and
whether their action has to be understood on a different basis is
another question, which we shall take up later.

If we try to discern the different possibilities as to how a series
of alleles may control a graded series of effects in an orderly way.
we have to realize that a series of reactions of different velocities
may act upon development in various ways, which have already
been discussed. If, for example, a case of pigmentation is
involved, a series of reactions of different velocities may lead to a
correspondingly graded series of effects:

1. By controlling the quantity of pigment formed. This ma}r
be accomplished: (a) by the amount of chromogen to be oxydized ;

(b) by the rate and degree of oxidation (or eventually reduction) ;

(c) by the control of the time of onset of one or the other or both
of these processes; (d) by the control of the time of the ending of
them; (<?) by controlling the time from which or during which the
organ or group of cells is ready to receive the pigment or its

THE MUTATED GENE 83

ingredients; (/) by controlling the threshold above which one or
the other of these variables is effective.

2. By controlling the method of diffusion of the ingredients of
pigmentation over the organ, which will show different grades of a
pigment pattern: (a) by controlling the time of onset and the time
of flow of the stream (see pattern, page 193) ; (6) by controlling
some process in the substratum which checks or enhances this
flow.

It is not claimed that these are all possibilities. But if one of
them is realized here and another there, certainly a number of
different types of reactions are imaginable which may lead to
graded effects by change of rate. This example for pigmentation
may easily be varied to fit cases of growth pattern, of production
of bristle-forming substance, or of whatever morphogenetic
processes capable of a linear quantitative variation are imaginable.

In a former chapter, we have already mentioned cases of
multiple allelomorphs which fulfill all the conditions of an
explanation by the concept of reactions of different velocities.
We point especially to the case of the vestigial wing in Drosophila,
where the different indications for this type of action of the genie
series were presented as follows:

1. The orderly arrangement of the effects in a quantitative
series which included also the compounds at the proper place.

2. The embryological proof that the actual process involved —
destruction of wing area — begins earlier and earlier in the
increasing series.

3. The fact that at the beginning of the series a typical over-
lapping with the normal takes place with a regular increase of
percentage in the ascending series.

4. The fact that the highest members of the series drift from
increasing in viability into lethality.

5. The fact that change of temperature shifts the effect within
the same series.

6. That effects of the same type and seriation may be produced
as phenocopies by temperature shocks or X rays.

7. That effects of modifying genes may produce the same shifts.

8. That certain phenomena of symmetry fit into the general
type of explanation (see page 229).

Another type of series already mentioned, which presented facts
that could be conceived only in terms of different rates of one

M PHYSIOLOGICAL GENETICS

reaction, were cases of pigmentation in guinea pigs or Drosophila
eyes. Numerous similar examples are found in genetic literature
(sec Stern, monograph, L930), which all lend themselves to the
same type of interpretation.

The next point to be considered is the action of pleiotropic
genes in such series. One and the same gene, in different multiple
allelomorphic conditions, acts upon many phenotypic characters,
and this may be done by changing the rate of one chain of
reactions which affects a developmental process, thus leading
automatically to manifold effects (see foregoing chapter). In
this case, the different characters affected ought to show corre-
sponding series of gradations, provided this is possible. (This
restriction means that the development of an individual trait
might be of an all-or-none type, which would exclude inter-
mediate steps.) It must be kept in mind, however, that such
an expected parallelism need not be of the simple type of simul-
taneous and parallel change of each character. Let us take an
example. A series of effects may be produced by an earlier and
earlier onset of some process, as in the vestigial case. As
embryonic development consists of a series of determinative
processes which narrow down the potentialities of parts of the
embryo, an earlier onset of a change will affect more potential-
ities. If we take the vestigial case, it is a fact that the imaginal
disk for wing formation contains also a thoracic bud and that
these twro areas are separated only during development. Let us
assume that a series of earlier and earlier changes occur in the
wing Anlage in the imaginal disk. They will affect only the wing
down to the stage at which determination of wing and thorax
Anlage occurs. At an even earlier stage, however, the still
equipotential Anlage is affected, and the resulting change occurs
in both wing and thorax. The latter effect then will appear in the
series only near its highest point. This example shows that in a
given case the same reaction might lead to effects in different
organs which might begin at different levels of the series. It will
be easy to imagine manifold variations of this one example.
The following diagram represents three such possibilities of
pleiotropic effects upon the three characters ABC, caused by a
series of allelomorphs n1, n2, etc., all resulting from one and the
same reaction, which hits different embryological processes at
different levels.

THE M I TA TED GENE

1 )| \',K\\I

85

Character

Phcnotypic expression grade 1 to 5 in the multiple allelo-
morphic series n1 to n5

n1

n2

n3

n4

?t5

A
B
C

5
5
5

1

4
4

3
3
3

2
2

1
1
1

A

5

4

3 2

1

B

3

2

1

c

2

1

A

5

4

3

2

1

B

2

1

C

1

in the case of an all-or-none response of character C:

A

5

4

3

2

1

B

5

4

3

2

1

C

5

5

5

In the section on pleiotropy, we have given an example, the
vestigial series, corresponding to the second diagram. Other
cases of such parallel effects, with or without shift between the
characters, are found in genetic literature. Here are a few
examples taken from Stern's monograph in table form:

Table 8

(Compiled from Stern's Monograph)

Dichaete Alleles in Drosophila

+
D<
D*
D

Bristles

Normal
Normal
Almost normal
Absent

Wings

Normal
Variable

Spread
Spread

86

PHYSIOLOGICAL GENETICS

Table 8.— (Continued)
Pharbitia Nil, the maple alleles (Imai)

Leaves

Corolla

Fertility

M

m
\1'

Normal
Maple

Willow-

Normal

Partly split
Split

Normal
Normal
Female sterile

Dro8ophila obscura,

white alleles (Lancefield)

Eye Color

Testis Color

Vitality

+
w"
w

Red

Eosin

White

Orange

Colorless

Colorless

Good

Medium
Poor

Drosophila willistoni deformed series (Lancefield)

Eye
Surface

Bristles

Hairs

Scutellum

Wings

Vena-
tion

+
d'
d

Normal

Rough

Rough

Normal
Almost normal
Bent

Normal

± Chubby

Chubby

Normal
± Normal
Round

Normal
± Spread
Spread

Normal
Normal
Defect

Drosophila melunogasler, lozenge eye

Smooth cornea

Fertility

+

Normal

Fertile

lz*

Increasing smoothness

Fertile

lzl

Increasing smoothness

Sterile

Iz*

Increasing smoothness

Sterile

k*

Increasing smoothness

Sterile

lz3

Increasing smoothness

Sterile

Rabbit, Russian pattern series (Kosswig)

Yellow pigment

Black pigment

Eye pigment

Cself

Present

Present

Present

Cch chinchilla

Almost absent

Present (?)

Present

Cd dilute chinch.

Almost absent

Reduced

Reduced

Cm sable

Almost absent

More reduced

Much reduced

Cr russian

Absent

Only at tips

Absent

C albino

Absent

Absent

Absent

THE M ( TA TED GENE 87

Many more variations are imaginable, all of which may be
explained by the action of a single process occurring at different
rates. There is an interesting fact to be mentioned in this
respect. We remember that most of the mutant characters of
Drosophila, including also rnultiple-allelomorphic series, could be
reproduced as phenocopies. Some such cases involved effects
which in the case of the actual mutant show pleiotropic phe-
nomena. Thus the series of the truncate genes in Drosophila
affects the wing in the series oblique — dumpy — truncate but
affects simultaneously the hairs on the thorax, which form a
vortex in different degrees. In Goldschmidt's experiments on
phenocopies, the types of the wing series up to dumpy could be
reproduced, but there wras no vortex. This shows that the hair
character has a different sensitive period to temperature shocks;
i.e., different details of developmental procedure are involved.
(See the facts regarding different sensitive periods in pleiotropic
characters of the flour moth, page 246.) This agrees with the
facts on pleiotropic multiple allelomorphs and their explanation.

There is another set of cases in which some of the characters
controlled by the allelomorphie series show parallel seriation
whereas others do not. To this belongs also the vestigial series
which we tabulated on page 80. In this table, wre omitted
another trait, found by Mohr (1932), viz., a shortening of the
second wing vein which is observable in the medium and lower
members of the series. This character does not follow the same
seriation as the others. If we use the same arrangement of the
£><7-scries as in Table 7 (page 80), the percentage incidence
of the character in question seems haphazard: 100, 2.1, 100, 0,
8.8, 73.3, 0, 2.6. If we look at the details of Mohr's data, how-
ever, we find a rule: The three high percentages are the com-
pounds in which the No-wing allele is involved. No-wing has
also other peculiarities, for it is semidominant to Wild type and is
homozygous lethal. From this one might draw the conclusion
that here some special feature is involved which we do not under-
stand yet but which has nothing to do writh the general chain of
reactions involved in the series.

The oldest and best analyzed example of this type is Wright's
analysis of the albino series in guinea pigs (1916, 1925). Here a
series of multiple alleles produces a progressive dilution of
pigmentation down to albinism. If the gene for black color is

ss

PHYSIOLOGICAL GENETICS

present, the effect ranges from black over sepia to white; if rod is

involved, it is shitted over yellow to white. Simultaneously, the
eye color is affected, ranging from black via red to pink. These
three series of effects, however, are not strictly parallel, as the
following table latter Wright, 1925) shows. In this table, the
colorimetric values for intensities of dark pigment are assembled:

Table 9

(After Wright)

Allele

With black

With yellow

Eye color

C

14

10.6

Black

CkCk

13.1

7.1

Black

CdCd

9.9

7.0

Black

C*Cr

13.1

0

Dark red

CCa

0

0

Pink (no pigment)

The table also shows that the seriation is parallel for yellow pig-
ment and eye color but not parallel for those in the blaek-
pigment series. Wright discusses these facts in a way that
ought to have settled the argument. There are two possibilities.
Either the black and yellow pigmentations are due to different
processes in development, in which case there is no reason why
any correlation ought to exist, and there is no problem at all.
Or the albino series affects primarily one and the same reaction
of pigment formation. In this case, the discrepancies of the
effect in the different series are due to developmental processes
subsequent to the effect of the albino factors. Wright thinks
(1916, 1925) that it is mainly a threshold problem, viz., that the
threshold for yellow is higher than that for black. "The red-
eyed dilutes may have a great deal of black in the fur but no
yellow7. The typically yellow parts of the tortoise-shell or the
agouti patterns are represented by pure white. Even albinos
develop black in the ears, nose, feet . . . but never a trace of
yellow." Similar facts are known for other rodents (see Wright,
1925). Wright shows that the available data reveal the same
situation in six mammals, pointing to a general rule regarding
threshold differences between black and yellow. We shall
return to the same facts in the chapter on dominance and also
in the chapter on the theory of the gene.

THE MUTATED GENE 89

This discussion touches the decisive points (if we leave out of
account here the reflections upon the nature of the genes involved).
(1) Facts of this type are expected if different processes are
involved. (2) If one process is involved, any special feature of a
developmental process that is in some way connected with the
main process will result in such phenomena as are discussed here.
Such developmental variation may depend on the control of a
threshold, but it may also be conceived in many different ways.
Most cases of this type have been reviewed in Stern's Monograph.
If we consider them from the standpoint of Wright's discussion,
there will be no difficulty in reducing the discrepancies of seriation
to such interferences as different thresholds, all-or-none effects
for certain processes, or changes of general situations of develop-
ment in an orderly way, such changes interfering only with the
differentiation of one organ (see the discussion in Goldschmidt,
1927c). Probably most frequent are those cases in which the
time relations of development (order of growth, etc.) are different
in different organs and the alleles affect some process at a definite
time. Let us assume A, B, C, D to be successive points in
development at which the alleles ml, m2, m3, m4 act. One action
determines the quantity of pigment in a linear series proportional
to the time of onset (A, B, C, D) of the reaction controlled by the
alleles. Another consequence of these reactions would be an
influence upon the growth of another organ, say speeding up
growth. If, now, the normal growth of this part should occur
at the times A, B, C, D in the following seriation: grow thin
breadth/growth in length, growth in length, growth in breadth,
then the alleles m1, m2, m3, m4 would produce the following seria-
tion of the length-breadth index in favor of length: 1- H , i.e.,

a different seriation from the pigment series. This model, which
might be varied in many ways (see Goldschmidt, 1932a, 1933c),
allows for all cases of not parallel pleiotropic action of multiple
alleles, without changing the general interpretation that such
genie series control different rates of one primary reaction.

Whether there are cases of multiple allelomorphs that act in
such an irregular way that no simple explanation is forthcoming
is another question. Many such cases are known, and they are
reviewed in Stern's monograph. Some of those may be brought
in line when the actual facts of development are known. Others
may turn out to belong to a different category of facts; and,

90 PHYSIOLOGICAL GENETICS

again, others may defy a simple explanation. This, however,
would not change the fact that most multiple-allelomorphic
scries give definite information regarding the action of the
respective genes upon development. More facts will be pre-
sented in other chapters.

C. Genes Controllincx Known Chemical Processes

The effect of a genie action may be a morphogenetic process of
any kind; it may also be the production of a physiological condi-
tion; and it may be the final production of a definite chemical
substance deposited somewhere in the body. In general terms,
we might assume that all genes act primarily in the same way.
The different types of end effects will therefore probably furnish
no different clues in regard to the primary reactions set in motion
by the genes. But there is a long chain of processes leading from
the gene to the final phenotypic expression of its activity. When
morphogenetic processes are involved, usually the members of the
chain that may be unraveled (see later chapter) cannot be defined
chemically but only in such general terms as growth substances,
evocators, hormones of differentiation. If, howrever, the end
of the chain is a definite chemical situation, there might be the
possibility of finding some members of the chain and of defining
actions of mutant genes in terms of definite chemical processes.

Strictly speaking, all the processes of deposition of pigments in
eyes, hairs, feathers, skin, many of which have already been
mentioned, belong here. But in most cases, these processes
have been studied only in a rather general way from the stand-
point of chemistry. In certain cases, it is known whether
melanin pigments are involved or pigments of the carotinoid type,
whether different pigments may be different grades of oxidation
or reduction of one substance, and similar rather general facts.
Such cases have been mentioned, and the fact was noted that the
phenotypic result might be dependent upon time or quantity of
formation of the chromogen (tyrosin, etc.) or similar quantitative
variants of the action of the catalyst (dopa, oxidase) or both.
More examples will be discussed later on. But as a rule the
chemical elements of these processes could not be clearly defined
(see page 91).

There are, however, a few actual chemical attacks on these
problems in animal genetics.

Til E Ml TA TED GENE 9 1

The first Mendelian studies of animal pigmentation tried to
explain facts in chemical terms. Cuenot (1902, 1911) assumed
that albinos do not contain chromogens but oxidizing enzymes.
Later, Gortner (1910-1913) showed that the color varieties of the
potato beetle contained different amounts of chromogen in the
elytra. Onslow (1915), on the other hand, found that in rabbits
the main differences are not due to the chromogen but to the
enzyme component. He actually found peroxidases in the skins
of gray, black, and brown rabbits, which oxidize tyrosin to
melanin in the presence of hydrogen peroxide. In all recessive
white skins or parts of skins, he found the peroxidase missing,
though the chromogen was available; and in the dominant white
English, an antityrosinase prevents melanin formation. The
same was found in the belly of light-bellied pigmented forms.
In yellows, no peroxidase was found.

Roller (1930) repeated the tests with the same results and
added that a pigmentation inhibiting extract from dominant
white does not act upon dominant black. (It ought to be added
that the relation found here cannot be generalized. In other
cases, genes for no or less pigment might control the chromogen
part of the reaction.) Wright (1916) used these facts — also
certain assumptions made by Little (1913) — to give a general
explanation of gene action in regard to coat color in chemical
terms. There is a basic color-producing enzyme (I) which
acting alone on chromogen produces a diffuse pigment which
appears yellow. The albino series of alleles produces this rather
unstable enzyme at different rates (see page 88). There is
another enzyme (also an oxidase) II, which stabilizes I and
makes it oxidize a chromogen into the coarser pigment granules of
dark pigmentation. I and II together set the threshold for
action of I at a lower level. The rate of II controls the admix-
ture of dark to yellow pigment in the different color types.
More details will be found later. Here belongs also the work of
Schultz (1935) on eye color of Drosophila which has been
discussed.

The same problem has been attacked by Graubard (1933) for
pigment races in Drosophila. He tried to compare the tyro-
sinase content of yellow, wild, black, and ebony flies. To do this
he devised methods for estimating the enzyme concentration,
ceteris paribus. The results were very strange. Different

02 PHYSIOLOGICAL GENETICS

results were obtained for larvae, pupae, and flies, also different
results for different treatments. For example, sometimes living
larvae yielded little or no enzyme, which was increased after
addition of chloroform or after grinding the tissues with sand.
Young flies did ao1 \ ield any enzyme at all. Graubard concludes
that the enzyme concentration may be of no importance at all
for melanin formation but that the decisive action rests with the
internal environment, which controls how much of the enzyme
may take part in the reaction. Another fact seems to corrobo-
rate this. Extracts prepared under the same conditions may
contain tyrosinase in the following diminishing order for the
races tested :

1. In pupae: yellow, wild, black, ebony.

2. In larvae: yellow, black, wild, ebony.

Graubard concludes, therefore, "that it is certain that Gold-
schmidt's picture of the gene-enzyme end-result relationship is
far too simplified." It seems doubtful whether this analysis lends
itself to reliable conclusions upon the genie action.

1. It is not necessarily the enzyme that is supposed to vary in
concentration; it might also be the chromogen.

2. The variable might be neither enzyme nor chromogen but
the time of onset or of ending of the oxidation process.

3. A measurement of the oxidase content of the whole animal
may be irrelevant for the happenings in the cells or organs that
deposit the pigment.

4. Methods that do not permit the discovery of any enzyme in
the young fly before darkening can hardly be called sufficiently
quantitative to permit of conclusions.

5. The assumed inhibiting action of the internal environment
may be identical with the selective distribution of the enzyme
to the regions of action.

In this connection, we must consider the work of Schmalfuss
and Werner (1926) who attacked the problem as chemists in
model experiments. They used as a chromogen an amino acid
( 1-/3-3, 4-dioxyphenyl-a-amino propionic acid) and as oxidase a
ferment from the blood of caterpillars. The blood was soaked
into strips of filter paper and dried. Chromogen was added to
these strips, and pigment formation occurred; the conditions of
pigmentation were then studied. It was first found that melanin
formation depends upon the pH. Acids interfere with the action

THE MUTATED GENE 93

of the oxidase; bases, except in high concentration, increase the
rapidity of melanin formation. There is no influence of light,
and a considerable one of temperature. This seems dependent
mostly upon the presence of a second catalyzer which is heatproof.
Systematically all types of substances were studied in regard to
their action upon melanin formation; some of them enhanced it,
while others interfered in different quantitative degrees (grades
from light grey through brown to black), which were measured.
Certain chemical rules were found, but these do not mean much
for the genetical analysis. The only conclusion that is warranted
is that, given chromogen and ferment, the amount and color of
pigment may depend upon a number of chemical conditions
which might be produced by other genes controlling them.

More suitable material is found in plants, because here the
basic chemistry of the pigments is thoroughly known after the
work of Wheldale, Willstaetter, Robinson, Karrer, and others.
Most of the work relating mutant genes to actual chemical
products has been done by Euler and his collaborators on the one
hand and the collaborators of J. B. S. Haldane (see 1935) on
the other, after the pioneer work of Miss Wheldale (later
M. W. Onslow) in early Mendelian days (see Onslow- Wheldale,
2d ed., 1925).

Euler and his collaborators studied in a long series of papers
(Euler et al., 1929-1935) the chemistry of chlorophyll-defective
mutants of barley, which had been genetically analyzed by
Nilson-Ehle and Hallquist. The results were checked with
similar mutations in other plants. There are the completely
white albino types of which seven genetically different ones were
tested; also the yellowish xantha and the pale alboxantha types.
The chemical differences between these types were analyzed. The
main results thus far are:

1. No chlorophyll whatsoever is present in the albino forms.
The heterozygotes, however, have normal chlorophyll content
and complete dominance.

2. The base of chlorophyll is the porphyrin-nucleus which
contains about 70 per cent of the mass of the molecule. The
white mutants produce only a small amount of porphyrin,
indicating that the disturbance occurs earlier than the porphyrin
synthesis. Carotin synthesis, which is chemically related, suffers
in the same way.

!H PHYSIOLOGICAL GENETICS

3. The content of katalase in the white mutants is reduced to
about one-third the normal amount. In the cotyledon, it is still
normal.

4. In two of the albino lines bill not in the others, also not in
colored lines, a methylated Indolyl base was found, indicating
specific chemical activities.

5. In albino forms, the plastids are defective; less so in the
xantha. This points to the possibility that the primary effect
of the mutant gene is upon the plastids. In this case, the gene
would not be concerned directly with chlorophyll synthesis, and
the facts would not help much lo understand the chemical side
of gene activity.

Before reporting upon the best analyzed case, flower color, we
shall mention a piece of work that is of a more physiological than
strictly chemical type, Brink's (1929a, b) studies on the waxy gene
of maize. This recessive gene produces among other effects the
presence of a specific type of reserve starch in the male gameto-
phyte, the pollen grain. As this gene segregates in the pollen
tetrad, the chain of reaction between the gene and its effect is no
longer than the time of the reduction division. Chemically the
two starches are very much alike as far as the basic sugar is con-
cerned, but Brink found that certain differences in reactibility
pointed to a structural difference. This seems to have to do
with the amylase, as extracts from waxy pollen give a lower
curve for glucose production than extracts from normal grains,
i.e., show a higher diastatic activity. Brink concludes from this
that the amylases in both cases may be different in kind and
therefore attributes to the mutant waxy gene the ability to
produce a somewhat different diastatic enzyme.

It is not necessary here to go deeply into the chemistry of
flower colors. Only such facts will be mentioned as seem to be
important from the standpoint of gene action. We shall follow
mainly the most recent wrork in the field by Miss Scott-Moncrieff
(1936). Flower color may be determined by a number of
agencies. If one pigment is involved, the color depends upon the
pigment in question and also frequently upon the pH of the
milieu. If more than one pigment is involved, different possi-
bilities are given as follows:

1. There might be a combination effect.

THE MUTATED GENE 95

2. There might be a background effect (just as in many
animal colors, e.g., the skin of a tree frog; of course no pigment
cells are involved in plants).

3. There might be the so-called copigment effect, which means
that a certain type of pigment is changed in color by the mere
presence of another pigment. The pigments involved are (a) the
plastid ones, carotin and xanthophyll; (6) the more important
sap pigments. The latter fall into two main groups, the antho-
xanthins and the anthocyanins. The xanthins have ivory to
yellow color and belong chemically to the group of flavones.
They may or may not occur as glucosides (containing a sugar
residue). They may produce independent color effects, act with
a background effect, act as copigments or quantitatively together
with the anthocyanins. Their genetic relation to the antho-
cyanins will be discussed later. As copigments they have a
bluing effect upon anthocyanin. The anthocyanin pigments are
sap-soluble glucosides which vary in color from scarlet through
violet and purple to blue. Generally, their structure is similar
to that of the xanthins, the fundamental differences being the
substitution of an O-atom at position 4 in the flavones, as shown
below.

HOr Y V< >OH HO

CI
O

<3

OH

OH

HO O HO

Flavone (luteolin) Anthocyan (pelargonidin)

They fall into three main groups: pelargonidin, cyanidin, and
delphinidin types, differing by the possession of 1, 2, or 30H-
groups on the side phenyl ring (the preceding formula has one
and is a pelargonidin). More than one of these pigments may
be present in a plant, and there are numerous chemical variations
such as methylation of one or more OH-groups, change's in the
sugar residue, and addition of an organic acid. These variations
and substitutions are caused by genetic factors. Most important
are the sugar residues situated at position 3 or 5. These might
be glucose, galactose, cellobiose, and others; <.g., pelargonin,

96

I'll YSIOLOGICAL GENETICS

cyanin, peonin, petunin, malvin, and delpbin arc all 3-5 glycosides
respectively of the pelargonidin, etc.

Mutations in flower color may be based on the different pos-
sible changes in regard to the anthocyanin formula, the copig-
ment, and the pH. The chemical change may be of many
different types. By way of illustration, we mention only a
single type, where the mutated gene prevents the methylation of
the complex, as shown below;

CI
O

CI

oh/V^s

O.CH3
^>OH

5/

O.CoHu06
OH
Malvidin-3-nionoside

CH3

O.CeHnOt
OH

Pelargonidin-3-monoside

The following extract from Scott-Moncrieff's table for Primula
sinensis indicates some of the different types of action of some of
the genes. Wild type is magenta by malvidin-3-monogalactaside
(primulin) copigmented by anthoxanthin (pH 5.3).

Table 10

(From Scott-Moncrieff)

Gene

Action on flower color

Appearance, pigmentation, and
pH of mutant

Modifies general anthocyanin

Pale coral or white pelargonidin

K

substrate to more intense 3.5
oxidized and methylated type

3-monoside pH 5.3

Produces anthoxanthin, copig-

Red; primulin, no anthoxanthin

B

menting and suppressing of
anthocyanin

pH 6.0

Produces localized acid pH in the

Blue; primulin and anthoxan-

R

petals and in the red corolla
tubes of p and g forms

thin (pH 6.0)

Produces specific anthocyanin-

Bright magenta; mixture of

Dt

pelargonidin 3-monoside

primulin and pelargonidin 3-
monoside pH 5.3

Inhibits anthocyanin in flower

Red stigma, dark center; pH

center and together with I) in-

5.3 red tubes pH 5.6

G

hibits the effect of R in petals.
Inhibits anthocyanin and the
effect of R in tubes

THE MUTATED GENE 97

In a similar way, a considerable number of plants with known
genetic basis have been analyzed by Wheldale (Onslow), Robin-
son, Lawrence, Hagiwara, Shriner and Anderson, Sando, Milner
and Sherman, and Scott-Moncrieff (1936). They have been
reviewed by this last investigator. From all the available
material Scott-Moncrieff derives some general information
regarding flower-color variation: In some cases, the action of the
different genes is independent and additive; in others, an inter-
action is found or a dependence of the effect upon the pH. The
following list covers the majority of known types of gene action:

A. Chemical Changes of Anthocyanin

1. Oxidation of the aglycone (without sugar residue) at 3' or 3' and 5'.

2. Oxidation and methylation at ibidem.

3. Methylation of the aglycone at ibidem.

4. Glycosidic change from 3 to 3-5 type.

5. Acylation.

B. Sap-pigment Production

1. Anthoxanthin and anthocyanin.

2. Yellow anthoxanthin background and interaction effects.

3. Ivory anthoxanthin copigment interaction and copigment effects.

4. General anthocyanin background and interaction effects.

5. Specific anthocyanin background and interaction effects.

C. Sap-pigment Regulation

1. General intensification.

2. General suppression.

3. Local intensification.

4. Local suppression.

D. Plastid Pigment

1. Production background effects.

2. Inhibition.

E. Localized Acid pH

Table 11 shows a few examples of such actions.

The problem is now to find the general rules that might be
deduced from such facts. We leave out of consideration here
the different interactions and localized effects, as these are prob-
lems of pattern formation and not directly concerned with genie
action; partly they are also problems of genie interaction to which
we shall return soon. One of the general results of such studies

MS

l'llYSloijHiKAL GENETICS

is thai the genes in question exercise a highly .specific action upon

such synthetic chemical processes as oxidation, met hylation,
addition of SUgar or acid residues at definite locations, and con-
trol of pll. in addition to the production of a more generalized
basic substance. Another point may be derived from the fact
that in many of the cases the Wild-type color is the most highly
oxidized one; mutant genes then prevent complete oxidation,
which again may work by controlling the rates of a decisive proc-
ess which might have to do with the specific catalyst involved in
the reaction. But, as Haldane (1935) remarks, it is very difficult
to draw definite conclusions upon the action of a gene from such
facts. "The gene which is responsible for the methylation of a
certain hydroxyl may be itself a diastase, may produce one, or
may be concerned with a reaction which liberates the necessary
energy for the action of an already present diastase." But, in
a general way, Haldane, like many others, assumes that the
action of the gene results, in most cases like those which were
mentioned, in the production of a specific catalyst.

Table 11

(After Scott- Moncrieff)

Examples of gene action

Varieties

Dominant Recessive Gene

I. Yellow plastids:

< 'keiranthua cheiri

II. Yellow anthoxanthin :
Antirrhinum majus

III. Ivory anthoxanthin copigment:

1'ri inula sinensis

IV. General anthocyanin:

Dahlia variabilis

V. Specific anthocyanin:

Pa-paver Rhoeas

VI. Oxidation of anthocyanin aglycone
Antirrhinum majus (3-rhamnogluko-

sides)

VII. Oxidation and met hylat ion of antho-
cyanin

Pelargonium zonale (3—5 dimonosides) . .

VIII. Local change in pH:

Primula sinensis

Yellow

Lemon

Y

Yellow

White

Y

Magenta

Red

B

Scarlet

Yellow

B

Scarlet

Pink

T

Cyanidin

Pelargonidin

Magenta

Malvidin
Pink

Magenta

Red

I'durijonidin
Salmon

Blue

X
R

THE MUTATED GENE 99

A special type of genie action controlling definite chemical
processes is, finally, the production of the stuffs that are respon-
sible for serological specifities. The bodies that make for the
blood-group differences belong here. A recent study by Abder-
halden (1936) has revealed another group. Rabbits injected
with proteins from other animals produce proteinases which
specifically split those proteins and which may be isolated from
the rabbits' serum. By aid of the specificity of these enzymes
Abderhalden demonstrated protein differences between different
linos of guinea pigs, distinguished by single genes. If this should
be confirmed, it might load to interesting conclusions.

6. THE GENE IN HETEROZYGOUS STATE

Since early Mendelian days, it has generally been assumed that
the phenomenon of dominance must one day lead to an insight
into the nature of the gene. Correspondingly, it was assumed
that dominance or recessiveness were attributes of individual
genes. Bateson, as is known, believed that dominance is the
presence and recessiveness the absence of the thing called a gene,
an idea that, however, was refuted at the outset by Correns'
(1899) discovery of dominance of two recessive genes over one
dominant in the triploid endosperm of maize. Later dominance
was attributed to different potencies of the genes (Davenport,
1908; Kellogg, 1908), which might change even individually.
( )nly slowly geneticists began to realize that dominance is not a
v" property of the gene but of the phenotype. It seems that
Tower (1910) was the first to report that dominance could be
changed at will in the same heterozygote of Leptinotarsa undecim-
lineata signaticollis by proper control of temperature and
moisture. At about this time, some geneticists realized that domi-
nance was not a property of the gene but something "floating,
capable of being shifted," as Goldschmidt expressed it in the
first edition of his textbook (1911a).

As far as I know, the first trial to express dominance in terms
of development was made by Goldschmidt (1916c, 1917c),
when he found that a case of so-called change of dominance dur-
ing development had to be explained by the respective velocities
of processes of progressing pigmentation. Dominance then was
controlled by certain time relations of processes or reactions
occurring during development which could be expressed in a

100

I'llYSloUHlKM. UENETICS

general way by the words: The faster reaction wins the race.
The further development of genetical knowledge, especially of

the interaction of genes, the action of so-called modifiers, led to
a general recognition of the idea that dominance is not a property
of the gene but a phenotypic result of the action of the gene in
connection with other gene-controlled developmental processes
or environmental actions. Wright (1925), in connection with
the fact that the Wild-type gene C is dominant over all members
of the albino series, designated such a process a threshold
effect, controlled by other genes. A generalized theory of
3 4 56789 10 II

Fig. 23. — Diagrammatic representation of dominance relations if the number
of cell divisions determines the phenotype, if and the effect of the heterozygote
is intermediate between that of the homozygotes. M, M\, J/2 are different sys-
tems of incidence of cell division. {From Goldschmidt, 1927. Phys. Th. d. Ver.)

dominance as a phenomenon of development was developed by
Goldschmidt (1927c), who explained the phenomenon as a neces-
sary consequence of the type of action of the gene in controlling
development. This theory is still the only one adaptable to all
facts and the only one that furnishes the necessary physiological
basis to all phylogenetic theories of dominance (see page 121).
The following case may serve as a model which might be adapted
easily to other cases by changing the respective variables. Let us
assume two allelomorphs controlling a character which might be
measured in quantitative terms, like the number of facets in an
insect eye. The dominant gene controls a reaction of a certain
velocity; the recessive allele, a reaction of lower velocity; and the
heterozygote, an intermediate reaction (see Fig. 23). The
reaction leads to a determining effect when the products have

THE MUTATED GENE 101

reached a definite concentration at the level M. For simplicity's

sake, we assume that the hereditary character in question is
the size of something, controlled by the number of cell divisions.
For our model we assume that the reaction in question is one that
stops further cell division. If the progress of cell division up to
this point is determined independently, the size of the organ
will be smaller the earlier the curve reaches the level M. If,
now, the cell divisions proceed in equal intervals, the hetero-
zygote Aa will be exactly intermediate between AA and aa.
Let us now suppose that the rate of cell division increases with
time as marked on line M. (the numbers indicate succeeding cell
divisions) and that the number of cell divisions alone decides the
size of the organ. In this case, in the heterozygote, cell division
ceases before six divisions have been completed, and therefore
the small size (.4.4) appears almost completely dominant.
(Stopping after five divisions characterizes A A.) If cell divi-
sions proceed faster at first and then slow up (line M2), almost
complete dominance of the large size (aa = 11 steps) results.
Dominance in this case is then conceived as a function of the
following variables: the velocities of reaction controlled by the
main gene, the threshold concentration at which this reaction
leads to a morphogenetic effect, the type of effect, the numerical
system of the independently determined process that is involved.
It is clear that this simple model allows for innumerable varia-
tions. To mention only one, the reaction A may stop or initiate
or slow down or speed up cell division or growth between the
divisions. It is clear that this type of model may be easily
changed to fit any imaginable gene-controlled process; and of
course other variables of a rate type, threshold type, or all-or-
none type may be easily added to describe the most different
cases.

Wright (1934d) has used the same type of reasoning but put
more emphasis, as Goldschmidt also originally did, on the rate
of reaction catalyzed by the genes in question. This means that,
though not neglecting the other variables of our model (Fig.
23), he emphasized specially the rate relations between the
products of dominant, heterozygous, and recessive genes, which
in our model have been assumed to be of a simple type. Wright's
deductions may therefore be added to that model as a further
analysis of one of its variables. Wright starts from the neces-

102 PHYSIOLOGICAL GENETICS

sary assumption that there is a chain of reactions between the

gene and the end product and that any condition that makes
one link in the chain act in the heterozygotes as it acts in the
homozygotes will result in dominance. He assumes that the
gene acts as a catalyst controlling the rate of production of some
substance (in harmony with Goldschmidt's basic assumptions)
by a chain of irreversible transformations, the intermediary sub-
stances being in a flux equilibrium. These reactions may be
treated as monomolecular and dependent upon the joint con-
cent rat ion of catalyst and substrate. The rate in this case
varies directly with the concentration of the catalyst, and the
curve expressing the relation of variation in the product with
variation in the catalyst is a hyperbole, asymptotic at its upper
limit. Therefore with increasing activity of the gene its hetero-
zygous effect ought to approach dominance. Another pos-
sibility is that a bimolecular reaction is involved. The general
consequences are shown to be the same, though a number of
new possibilities appear, e.g., one part of the reaction setting a
threshold for another. A third possibility is assumed, viz.,
that the rate depends not separately on concentration of sub-
strate and enzyme but on their compound. Also, in this case,
a similar result will follow, viz., a relation between the difference
in the quantity of the catalyst and the amount of dominance.
Without going into further details of Wright's calculations, we
realize that in his view again dominance is dependent upon the
same elementary gene-controlled processes as are involved in the
model (Fig. 23), emphasizing more the behavior of the primary
chain of reactions. Thus we see that dominance is best con-
ceived as a consequence of the quantitative aspects of genie
action and their quantitatively fixed interplay in time and
space during development. If this is correct, dominance, vice
versa, furnishes important facts toward the understanding of
genie action. To these facts we shall turn now.

A. The Influence of the Environment upon Dominance

If dominance is controlled by a developmental system, as
indicated above, it ought to be possible to shift the phenotype of
the heterozygote by environmental action in the same way as
the mutant phenotype may be secured as a phenocopy. This is
especially to be expected when the mutant phenotype appears

THE MUTATED GENE 103

as a simple quantitative deviation from Wild type toward insuffi-
cient production of something needed for perfect development.
Muller (1932a) has called this type of mutant gene hypomorphic.
It is the most frequent type of recessive mutation. We might
distinguish two possibilities of such actions. In cases of incom-
plete dominance, the degree of intermediacy might be shifted
experimentally. Or in cases of complete dominance, this might
be broken. A priori, the former type of effect would be expected
to be of general occurrence, being, after all, nothing but a special
case of the general type of modification. It is, however, to be
expected that the second type of shift is more difficult to accom-
plish, because complete dominance requires certain threshold
conditions (see page 117), with the result that even a considerable
shift does not carry the type below the threshold.

Examples for the first group may again be taken from the
much studied case of vestigial wings in Drosophila. We recall
that the vg-w'mg may be changed by temperature action into the
phenotypes of all the other alleles. M. H. Harnly and M. L.
Harnly (1936) discovered a new allele pennant which has normal
or slightly notched wings and reacts strongly to temperature
changes in different ways. 1 The heterozygotes with v g may show
very different types according to the temperature at which they
are raised, fluctuating from vo-strap type at 26 to almost normal
at 32°. Here, then, temperature shifts the heterozygote exactly
as it shifts the homozygous vg-wing. The explanation in terms
of rates is obvious. It must be emphasized that the vg/ ' + -
heterozygote is usually Wild type in all temperatures (see
page 117).

At this point we have to mention a phenomenon that resembles
a shift of dominance and has actually been confused with it
(see also preceding footnote concerning the same error by
Harnly). Hersh and Ward (1932), among others, compared
wing length and area in homozygous -f - flies and the hetero-

1 M. H. and M. L. Harnly base their results upon measurements of wing
areas. This leads to erroneous results, the details of which will therefore
not be reported. His data show that the pennant wings are affected by
temperature considerably in general size, i.e., in cell growth after differentia-
tion. The measurements then lump together this influence upon secondary
growth with the effect upon the scalloping of the wing, which is a different
process. Goldschmidt's results in this respect could not have been known
to Harnly.

lot PHYSIOLOGICAL GENETICS

zygous +/vg- flies at different temperatures. In both cases,
the wins area decreases with rising temperature. This they
treated as a dominance effed and calculated even dominance
coefficients. The assumption, of course, was that vestigial
produces a small area. As we know now from Goldschmidt's
work, the vg-eSeci has nothing to do with wing growth but with
destruction of parts of the wing. A wing without notches is
therefore completely dominant. If temperature affects the wing
area in either homozygotes or heterozygotes, the ygr-gene is not
involved at all but a growth process, primary or secondary (prob-
ably only secondary) cell growth. The results therefore have no
bearing upon the action of the y^-gene or its normal allelomorph
but only upon the problem of whether or not the yo-heterozygote
reacts differently from the Wild type in regard to growth under
different temperatures. The small difference actually found may
have many causes, which might or might not allow conclusions
upon genie action.

The most complete and most thoroughly analyzed data have
been presented by Hersh (1930a) for the Bar-eye mutants of
Drosophila. Here dominance of the Wild type is not complete
(the Drosophila workers call Bar, therefore, a dominant, a rather
inexact but useful way of labeling), and therefore a shift of
dominance is to be expected. The number of facets in Wild-
type eyes and the eyes of the mutant series is a function of tem-
perature and may therefore be expressed as a curve for the
range of this variable. Hersh (19306) calls such a curve a
thcrmophene. If such thermophenes for Wild type, mutant, and
heterozygote were compared, they might be parallel. This
would indicate that a general process of growth is involved but
not the phenomenon of dominance. If a measure for dominance
should be introduced, it would remain inconstant through the
range of temperatures. If, however, the curve for the hetero-
zygote should not parallel that for the homozygotes, conclusions
upon dominance might be drawn. In this case, an eventual
coefficient for dominance would change over the temperature
range. Hersh uses Zeleny's (1920) formula for calculating a
coefficient of dominance:

A A - AAX
Co = AT^A^ X 10°

THE MUTATED GENE

105

in which AA and AiA] are the obtained quantities for the char-
acter in the two homozygous and AA\ in the heterozygote.
Instead of using directly the facet counts, he applies a dynamic
view by considering the instantaneous addition of facets. He
had found before that the rate of change in mean facet number
at any temperature is proportional to the mean facet number at
that temperature, i.e., an exponential relation of the formula
y = aert, in which r is the relative rate of change, t the tempera-
ture (centigrade), e the base of the natural logarithms, and a
a constant. The instantaneous change in facet number at a
given temperature is then expressed by the first derivative:

dy
dt

yr

This value he calculates for the different facet numbers at dif-
ferent temperatures and different genetic constitutions, and from
these values, i.e., the instantaneous change in facet number
instead of the direct facet counts, he calculates the coefficient of
dominance according to Zeleny's formula. The following table
contains the results for the heterozygotes between Full eye,
Bar, and Ultrabar for reciprocal hybrids. The second line of
headings designates the mothers of the heterozygotes (recipro-
cal crosses).

Table 12

(From Hersh)

rpo

Full over Bar

Ultrabar

over Full

Ultrabar

over Bar

Bar

Full

Ultrabar

Full

Ultrabar

Bar

15

81.58

61.69

78.20

55.66

76.74

55.51

18

74.11

60.64

82.26

66.80

76.01

58.07

21.5

68.18

59.00

85.93

76.32

74.84

61.19

25.5

62.81

56.62

89.13

83.96

73.33

65.08

27.5

60 47

55.38

90 46

86.82

72.48

67.44

29.5

58.29

54.10

91.60

89.25

71.42

70.00

This shows that the degree of dominance in regard to the
number of facets formed at a given moment changes regularly:
it decreases in the -f-/Bar case with increase of temperature from
dominance of Wild type; it increases for dominance of Ultrabar

106 PHYSIOLOGICAL GENETICS

over Wild; it decreases or increases in reciprocal crosses for
dominance of Ultrabar over Bar. Hersh draws the general
conclusion from these tacts that in its capacity as a modifier of
growth during the course of development, the more dominanl
gene has relatively greater or less effect as the growth changes
are affected by the temperature. He then seems to think thai
the activity or the potency of the two allelomorphic genes (which
are thus complements! to each other) has been measured. I
should prefer to assume that the temperature coefficients of the
rates of the developmental processes controlling the beginning
or end, etc. (see Fig. 23), of the facet forming (or iacet-A nlagc
destroying) process have been measured or of such processes as
control the underlying numerical system of cell divisions or of
both. We shall again consider this same set of facts.

Hersh (19346, c) returned later to the same problem in connec-
tion with the remarkable difference of temperature action upon
Bar and Infrabar (Luce, see page 143). It will be shown in
detail (page 143) that in experiments involving Bar, Infrabar,
and the heterozygote, Bar is dominant at all temperatures. As
Bar produces fewer facets with increasing temperature, but
Infrabar more facets, the dominance of the heterozygote must
in cicase in favor of Infrabar with a decrease of temperature.
Near 17° this leads actually to a reversal of dominance, as
illustrated in Fig. 29 (page 143). From this the conclusion may
be drawn "that any theory of dominance . . . based on the
quantitative serial order of the1 phenotypic effects at 25° would be
inconsistent with a theory based on the serial order of the data
at 17° or below." Such a conclusion does not, however, apply to
the type of theory of dominance represented in Fig. 23; here only
the difference of reactivity to temperature between Bar and
Infrabar would have to be added as a further variable Involved.

In the chapter on phenocopies, we mentioned that Friesen
(1936) succeeded in producing the same types of phenocopies in
Drosophila by the action of X rays as Goldschmidt (1929a,
1935a) had done by temperature shocks. Both shocks are sup-
posed to act upon rates of reactions connected with morphogene-
sis. As we do not doubt that the shift of dominance is primarily
also concerned with rates of developmental processes, we might
expect that X rays will also be able to perform a shift in
dominance.

THE MC TAT ED GENE 107

Temperature shocks, finally, have also been found to influence
dominance even in cases of complete dominance, where ordinary
temperature effects are powerless. The heterozygous ebony
flies (Drosophila) will show a certain amount of the recessive
melanism if the larvae have been treated with temperature
shocks (36 to 37°); in heterozygous vestigial flies, the same
treatment might produce erect scutellar bristles which is one
of the recessive traits of the vestigial fly (Goldschmidt, 19356).

B. DOMINIGENES

For a long time it has been known that dominance may be
affected by the presence of other genes acting as modifiers for
dominance (conditioned dominance.) These genes might have a
visible mutant effect besides this modifying action, and they
might also be without any other visible effect. When it is
impossible to separate such genes, they are sometimes spoken
of as the general genie environment (Tchetverikoff). As a short
term for such modifying genes, Goldschmidt (19356) has pro-
posed the word dominigenes. Bridges (1913) found that colored
eyes in Drosophila were incompletely dominant to white in the
simultaneous presence of the genes vermilion and pink. Lance-
field (1918a) found in Drosophila a third-chromosome gene
(semiforked) which shifts the dominance of Wild over forked
toward the forked side. Landauer (1933) found a dominigene
for the frizzle character of fowl, and Dunn and Landauer (1934)
showed that dominigenes exist that shift the type of the hetero-
zygote for Rumpless (a dominant mutation) toward the normal
type. Lebedeff (1932) found for D. virilis that the recessive
gene ruffled acted as a dominant in the presence of another gene
rounded. Crew and Lamy (1932) found a similar case for
D. obscura: the gene purple acts as a dominigene upon the
hetero zygote -j- /vermilion, vermilion being recessive. These
last-named authors — the author believes the only ones — try to
give for this action an interpretation, to which however we are
unable to agree. They believe that genes in other loci may have
exactly the same action as the pair of genes studied. Their
presence might therefore interfere with dominance by an act of
quasi allelomorphism. As in such a case three Nonwild-type
genes would be present in the heterozygotes, the effect might
become visible in the form of what looks like change of dominance

I OS PHYSIOLOGICAL GENETICS

(Crew-Lamy, 1932). Another case in rodents is described by
Burrows (1934). An unknown number of dominant dominigenes
makes the heterozygote of Agouti and black intermediate instead
of Agouti.

Such eases, according to Schultz (1935), do not seem to be rare
in Drosophila. He found in 37 combinations of a heterozygote
of one eye-color mutant with another homozygous mutant five
cases of conditioned dominance, as the phenomenon is sometimes
called, without being able to establish a rule connecting this
phenomenon with the process of pigmentation, which we described
on page 33. A few similar facts have also been known for
rodents where recessive albinism and spotting may become
incompletely recessive in the presence of other genes (Wright,
1927; Cuenot, 1911), and Snell (1931) described a case that he
classifies as a change of dominance in an individual, in which
at first dilution appeared dominant, whereas later the normal
recessiveness was restored. The case is, however, not completely
clear. Kikkawa (1934) described for Drosophila virilis an
enhancing action of the gene confluent upon the dominance of
plexus. But here another wing-venation gene is involved, and
therefore the case might also be regarded from a different angle.

In some other cases, dominigenes were found after crossing:
Harland (1932, 1934) found that a recessive mutation of cotton
behaved as an incomplete recessive if crossed with different varie-
ties which are supposed to possess dominigenes for this character.

A comparable case had been known for Drosophila. Morgan
and Sturtevant (1929) found that in hybrids of D. melanog aster
and simulans the genes Bar and Lobe are no longer dominant.
Furthermore, the expression of Delta from melanogaster is less
extreme, but the expression of Delta from D. simulans is more
extreme in the hybrid than in the pure species.

A parallel case in mammals has been described by Green
(1936) for the short-tailed mice, the genetics of which we men-
tioned before. He backcrossed heterozygotes of this dominant
mutant to a normal strain; the resulting heterozygotes showed
less dominance of the abnormality. The same heterozygotes
were outcrossed to a different species {baetrianus) , and still more
normal tails resulted. In the first case, the introduction of one,
in the second of more than one, dominigene is assumed. To the
same class of phenomena belongs also Fisher's (1935) recent

THE MUTATED GENE 109

work. He tested (from the standpoint of his phylogenetic
theory, see page 121) the dominance of the dominant mutations
in fowl. He found that after repeated backcrossing to the wild
jungle fowl such genes were not dominant any more, probably
because by this time some dominigenes, necessary for dominance,
had been removed.

In some cases, the existence of dominigenes was proved by
selection. Timofeeff (1925, 1934) selected for modifiers of two
recessive abnormalities of the venation in Drosophila funebris
and succeeded in establishing lines with incomplete recessiveness.
A similar result was obtained by Stanley (1935) for the recessive
vestigial wing. Finally, some cases ought to be mentioned that
are more complicated because chromosome abnormalities are
involved. Mohr (1929a) showed that the effect of the dominant
deficiency Gull in Drosophila is lessened by the presence of the
gene dachsous. Of course, a deficiency cannot strictly be com-
pared with a heterozygote; this point will be discussed in another
chapter. A similar observation is to be made, when it is reported
that the addition of a duplicated fragment of a Wild-type
chromosome in Drosophila to the recessive line does not result
in a completely wild type (Dobzhansky and Sturtevant, 1932).
Many possibilities are given to explain such cases within our
general model. A very different point of view, the so-called
position effect, will be discussed later in connection with the same
facts.

The most thorough analysis has been made thus far by Gold-
schmidt (1935-1937) for the recessive character vestigial wings
in Drosophila. Normal wings are completely dominant over
vestigial. It was possible to isolate three dominigenes which
shift the dominance toward the vg-type if all three are present.
These are one sex-linked recessive allele of the cut gene and two
autosomal dominants in the second and third chromosomes.
If only the autosomal dominants are present in the vg/ + hetero-
zygote, a small dominance effect is produced; i.e., about 1.1 per
cent of the individuals are nicked. The cut allele in homozygous
condition affects similarly about 0.5 per cent individuals. If,
however, all three dominigenes are present, all vo/+ heterozygotes
are scalloped. The autosomal dominigenes in this case have an
additive effect : for the degree of scalloping, namely nick to notched,
(see page 57) increases from AB/ab to AB/AB. A gene was also

110

PHYSIOLOGICAL GENETICS

found that counteracts the effect of these dominigenes and is

probably identical with the gene for black; in addition, the
quantitative effecl is different in both sexes.

In this case, further facts could be found which allow the draw-
ing of conclusions upon the phenomenon of dominance (Gold-
schmidt and Hoener, 1937). In a former chapter, we reported
(page 56) that the series of conditions of scalloping that connect
the normal with the vestigial wing may be produced by a series

1 vq '
"AW

w

FSL..Qp-D \no

QQYn

vgf

"9

-i 1 r

m

"9

Q* D .
Vw ni
Nw

Nw>

n

.BQp
no

™D-

yi x ix vni m

VI v

IF M JT I

Nw

wD

vg

no. ■ ni

ni

<* no

'D

Wild type

Fig. 24. — Graphic representation of the action of the dominigenes upon the
phenotype of heterozygotes and homozygotes of the vg series. (From Gold-
schmidt- Hoener, 1937, Univ. Cat. Publ. Zool. 41, Fig. 1.)

of multiple alleles of vg and the respective compounds. A similar
series may be obtained by temperature action upon the vg
homozygote (see page 12) and again by temperature or X-ray
shock upon the Wild type in the phenocopic experiments (see
page 8). It was further reported that the effect of these
genes is the production of a degeneration of already formed
wing tissue according to a definite pattern and that the degree
of scalloping is proportional to the time of onset of the degenera-
tive process (page 26). These facts have already been used
to derive certain notions regarding the action of the gene, and
we shall meet them again in the chapter on pattern (see page 229).

THE MUTATED GENE 111

We may interpret them by means of a simple diagram (Fig. 24),
based upon the following considerations. Different genie
combinations produce different percentages of scalloped and
normal, in orderly progression. As was pointed out before
(page 67), this must mean that there is a threshold beyond
which individuals are normal and a fluctuation that shifts part
of the population across this threshold. This threshold line
is found in the diagram as a definite time in development, beyond
which the wing remains normal, and the fluctuation of each
population is represented by the base line of the statistical curve
of variability. To the left of the threshold follow the different
degrees of scalloping from I = nicked to XI = no wing. The
unbroken lines represent the range of fluctuation for populations
of the different alleles and compounds (cf. Fig. 20, page 59).
In the experiments to be reported, the dominigenes were added
to all these genie combinations, and the broken lines represent
these combinations. We see here that the phenotype of all
the heterozygotes, homozygotes, and compounds up to homo-
zygous notched are shifted in a perfectly parallel way toward less
dominance of ± normal, i.e., to the left side of the diagram. This
regular behavior in the action of the dominigenes is found up to
class VI of scalloping but not beyond. Class VI is the maximum
scalloping possible if degeneration sets in after pupation, and
therefore the dominigenes in this case act only upon processes
after pupation, i.e., after evagination of the imaginal disks. The
important point is now that the dominance-shifting action
of the dominigenes works in a consistent quantitative way upon
different heterozygotes as wrell as homozygotes of lower alleles:
the shift is then not a shift of dominance but a shift in a process
that is responsible for dominance as well as for the phenotype
of the allelic series and their compounds (also for the phenocopic
effect). Since it was stated earlier that these phenotypes are. the
product of different rates of one and the same reaction (viz.,
the production of growth substances, or of lytic substances.
see page 59), it follows that dominance also is a subphenomenon
of the same system of reactions of different speed: dominigenes
change this speed or one of the concomitant processes controlling
threshold, etc. (see model Fig. 23). with the same quantitative
effect upon the result as different alleles or temperature shocks
produce. Only a system of the type of our modc^l (Fig. 23) will

112 PHYSIOLOGICAL GENETICS

account for such a combination of consistently quantitative
effects from different causes.

We shall return to this case in the chapter on dominance in
allelomorphic series.

C. Change of Dominance

In early Mendelian literature, occasionally changes of domi-
nance within an individual were reported; e.g., in the cross of
red and yellow snails (Helix), the younger shells are yellow and
become red later (Lang, 1908). Many similar cases have later
found a different explanation or have not been studied suffi-
ciently. But there are some cases known that actually may be
described as a change of dominance. Goldschmidt (1917d,
19206, 19246) described such a case for the pattern of the cater-
pillars of Lymantria dispar. There is a race with bright-yellow
epidermal markings which are visible through the transparent
euticula and remain constant through all instars. There is
another race without these markings, except in very young
stages, in which the euticula is incrusted with dark pigment.
In Fi, the young caterpillars are light but, in the course of
development, become darker and are finally all pigmented.
This might be described as a change of dominance from light
to dark. What actually happens is that the cuticular pigment in
the hybrid is at first not present but slowly accumulates through
the instars, thus covering the underlying bright markings.
A comparison between the two races and the hybrid then shows
that the gene for pigmentation of the dark race controls a rate
of pigment production of such velocity that even the young
caterpillar is deeply pigmented. The allele present in the light
race controls a pigmentation process that is so slow that usually
it does not produce visible pigment. The heterozygous gene
(and also different alleles of the gene for pigmentation) controls
a rate of pigment production that is intermediate; therefore the
pigment covers the bright pattern only in later stages of develop-
ment, which then explains the phenomenon of change of domi-
nance in terms of rates of deposition of pigment.

Another case that lends itself to a similar explanation has
been found by Honing (1923). He found a mutation of the Deli
tobacco called deformis which among other characteristics has
irregular narrow leaves endowed with excrescences on the lower

THE MC TA TED GENE 1 1 3

surface. The Deli, the deformis, and the hybrid plants begin
development with normal leaves; deformis forms early the char-
acteristic excrescences of the abnormal leaves. The hybrid
begins a few weeks later with the same process, and the Deli
race never does or does only later under certain conditions.
This, again, looks like a change of dominance, but it is easily
explained by the causation of chains of reactions of different
velocities leading to the incidence of the abnormality and having
three ascending velocities in the normal, the hybrid, and the
abnormal race. Honing also derives this explanation as a parallel
to the Lymantria case.

Later, Honing (1928) reported another instance. There are
races of tobacco that need light for germination and races that
do not. In young seeds of hybrids between such races, the need
for light is dominant; in old seeds, recessive.

Another comparable case has been described by Ferwerda
(1928) in the flour beetle (mealworm), Tenebrio molitor. Here a
pair of allelomorphs control orange and brown color through all
stages of life. In the hybrid, however, the larva and the pupa
are more orange; the imago, brown. It is easy to derive an
explanation on the basis of reaction velocities, considering the
facts regarding melanin formation (discussed in other chapters).
As a matter of fact, the case offers a close parallel to the develop-
ment of larval pigments in Lymantria. Ferwerda derives such
an explanation and constructs reaction curves for illustration.
If the product of reaction is the amount of chromogen produced,
the explanation derived for the pigmentation of Lymantria
caterpillars also fits this case completely under the assumption
that more chromogen is equivalent to deeper color. Other
interpretations of the same general type are easily imagined.

D. Dominance and Chemistry of Products of Gene Action

A few facts of a chemical nature are available that may one day
throw further light on the system of reactions within which the
quantitative shift is supposed to account for the facts of domi-
nance. We have already reported Schmalfuss' chemical models
of the oxidation of chromogens. It was shown that a number
of different factors influence the degree of this reaction, and it is
conceivable that, as far as melanin formation is concerned, the
model might be useful, showing that everything external,

114 PHYSIOLOGICAL GENETICS

internal, genetic, controlling these different conditions may act

upon dominance. But no actual work on the chemistry of
heterozygotes has been done.

There arc also some data from Scott- Moncrieff's work on
anthocyanins which arc relevant. Though this author is not
inclined to lay down definite rules in regard to dominance,
she points out some uniformity in behavior: plastid pigment,
copigment, anthoxanthin, and anthocyanin are dominant to
their absence. More oxidized and more methylated pigment
is dominant over less oxidized or methylated. Diglycosidic
and acylated anthocyanins are dominant over monoglycosidic
and normal anthocyanin. More acid sap is dominant over less
acid. It is obvious that such results may be interpreted in a
general way in terms of reaction velocity, the phenomenon that
we have always found in some way underlying the phenomena
of dominance.

E. Dominance in Multiple Allelomorphic Series

A considerable body of facts regarding dominance have been
derived from the study of multiple allelomorphic series, and
some of them have already been reported. A number of authors
have realized that such facts may have an important bearing
upon the understanding of gene action, viz., Goldschmidt, 1917-
1927; Zeleny, 1920; Wright, 1925; Hersh, 1930a; Dobzhansky,
1930; Muller, 1932a; and Goldschmidt, 1937. The pertinent
facts up to 1930 have been brought together and analyzed
critically in Stern's monograph (1930). We shall recount first
the types of facts, closely following the grouping employed by
Stern, as well as his examples, and then try to analyze them.

There is first the group of multiple alleles with a simple quanti-
tative sedation of the effects, not considering now the cases of
pleiotropic action of the same genes. Here the following types
of behavior in regard to dominance are found:

1. Each higher allele is dominant over the lower one. Exam-
ple: flower color of Lathyrus: RE self-colored dominant over
rV; spotted, over rr white.

2. All alleles are intermediate in the compound. Example:
the series of genes for eye-form Lobe in Drosophila +, L, L3, L2;
the pigment series of Lymantria caterpillars.

THE M U TA TED GENE 115

3. The most frequently found cases show complete dominance
of the highest member of the series over all others and inter-
mediate behavior of all compounds not involving the highest
member. To this group belong many of the best known series
in animals and plants, e.g., the white-eye series of Drosophila
(wild, coral, blood, cherry, apricot, eosin, ivory, tinged, buff,
ecru, white).

4. The theoretically most important is the case in which the
situation is as in the last-mentioned group, but in addition the
highest member is not dominant over the lowest ones. To this
important group belong also the cases in which the lowest mem-
bers are more or less lethal in homozygous condition. Examples
of this type will be discussed in detail later.

5. Special cases requiring independent consideration, e.g.,
the pattern of silkworm larvae: Ps striated, pm dark brown,
P normal, p unpigmented. From pm to p the darker alleles are
dominant over the lighter ones. P is dominant over the two
lower alleles but only incompletely so over the highest pm. But
of course here no strict sedation is involved, as striped is a new
character in the series.

There is a second set of facts which become visible when
different effects of the same series upon different traits are
involved (pleiotropy). Here the following possibilities are
realized:

1. The different effects may be arranged in parallel series, and
within these series dominance is also parallel, e.g., the albino
series of rabbits and mice affects in a parallel seriation black
pigmentation, yellow pigmentation, and eye color; and in
each the highest member is dominant, and the other com-
pounds intermediate. But (as a threshold effect, see page 85)
this does not mean that the highest member is always only the
first one.

2. Within each series of pleiotropic effects, dominance behaves
in an orderly way; but the series are not parallel in the different
traits affected; e.g., the aleuron-color and pericarp-color series in
maize (R and P) affect the color of aleuron, vegetative parts,
anthers, and pericarp. Everywhere deeper color is dominant
over more dilute. But if the series of alleles is arranged for
aleuron from dark to light, the effect upon the three other
characters is not parallel; they have each their own sedations.

116 PHYSIOLOGICAL GENETICS

If, then, dominance is considered for all effects of one allele,
it acts simultaneously as a dominant on one character and as a

recessive upon anot her. An example from animals is the truncate
series in Drosophila with consistent but not parallel actions upon
wings and thoracic hair. In such cases, a special phenomenon
may occur, if some of the alleles act only upon one of the traits.
In the truncate series, the high allele Tv0 does not affect the wings
but the hair. The intermediate allele T° affects the wings but
not the hair. In the wing series, Tvo is dominant over T°; in
the hair series, T° is dominant over Tv0. The compound of both
then produces a normal phenotype. There are finally the cases
in which the action of multiple allelomorphs does not appear in
a quantitative series, e.g., the spineless and aristapedia alleles
of Drosophila, the first affecting the bristles; the second, a
homoeosis of the antenna. Here rather independent behavior
of dominance in regard to the different traits occurs. A special
type may be found if, as in the scute series of Drosophila,
the alleles affect patterns of the same structure; each allele
prevents the formation of definite bristles. In this case, the
presence of the individual bristle in the pattern is always domi-
nant over its absence, from which it follows that the compounds
are always, more like the Wild type than the effect of both ingredi-
ent genes would be. And if two alleles affect reciprocal parts
of the bristle pattern, their compound must be Wild type. But
such cases, though important in other respects, do not furnish
special information for the problem of dominance.

We have now to find out whether such facts regarding dominance
in multiple allelomorphic series furnish information on the mean-
ing of dominance and therefore throw light upon the problem
of the action of the gene. To do this we shall discuss first such
cases as, according to facts already reported, permit one to draw
definite conclusions. The best case in this respect is made out for
the vestigial series. The dominance relations in this case are
of the type as reported under 4 in the foregoing enumeration.
Wild type is dominant over all alleles except the lowest member
of the series, No-wing, which, in addition, is lethal in a homozy-
gous condition. All the other compounds are intermediate and
fit into their proper places between the homozygous alleles,
as may be seen from Table 7 (page 80) and also from Fig. 24.

THE M I 'TA TED GENE 1 1 7

In this case, we have important information on some decisive
points which makes possible a definite picl ure of the phenomenon.
1. There are first the facts concerning the lower members of the
series. As reported above and discussed from different angles,
these lower members produce their effects in a definite way.
The lowest member nicked produces no visible effect in homo-
zygous condition, according to Mohr. (In Goldschmidt's lines,
there was a very small percentage of nicked individuals. The
same shift as compared with Mohr's lines applies to other com-
binations which explains the small difference between the data
in table 7, page 80, and those used for construction of Fig.
24. The difference is probably due to different modifiers.)
The higher allele notched affects only 42.5 per cent individuals,
and it is only with still higher alleles that 100 per cent effect
is reached. If we represent these combinations and those com-
pounds that range between them (see pages 110 and 111), as is
done in Fig. 24, the effect of the allele is considered statistically
as a curve of variation, a more or less large part of which is
situated beyond the threshold for Wild type. (See also page
68, the first case of this type in Lymantria.) If we follow these
curves from no/no with 100 per cent scalloping to ni/ni with only
a few per cent scalloping, it is obvious that the curves for lower
combinations, all within Wild type, ought to show a continuous
shift to the right, as represented in Fig. 24. This is of course
only an extrapolation; but it can be proved correct. In Fig. 24
are also, represented as broken curves the actions of the same
genes and compounds, with the addition of the dominigenes
(see page 110). These shift the curves of the phenotypes to the
left, and they do it, as the figure shows, in a way exactly parallel
for the different combinations. Thus, the first heterozygote
m/+ is shifted a little beyond the threshold, and all the higher
compounds correspondingly more. This shows that there is
actually within the Wild type a corresponding arrangement of
more or less Wild-type individuals — degrees of Hyperwild, we
might say, which thus far could not be distinguished phenotypi-
cally. To this series would also belong at their proper place
the normal vg/-\- heterozygote (to be put between ni/ni and
no/ '+) and the dominigene action upon it, which in Fig. 24 have
been put into another group for certain reasons.

IIS PHYSIOLOGICAL GENETICS

2. As just mentioned, there is a similar phenomenon at the
other end of the scries: No-wing/ + 18 not of Wild type but shows
a certain percentage of scalloped individuals (1.3 per cent in
Mohr's case, more in (Joldschmidt-Hoener's counts). As Fig. 24
shows, this is shifted by the dominigenes to 100 per cent. In the
usual language, this means that the extreme allele No-wing is
somewhat dominant. If viewed in connection with the fore-
going facts, it has a very different meaning, as it fits now into
an orderly series. We know already (see page 28) that the
scalloping effect is produced by destruction of already existing
wing area, presumably by insufficiency of some growth substance
or by accumulation of some lytic stuff. Let us now assume,
adapting a deliberation of Mohr (1932) to the new body of facts,
that the amount of growth substance — to use only this possi-
bility; it could be just as well presented for lysis — produced by
the alleles could be measured and that the threshold for the
production of Wild type would be 40. (This method of analysis
goes back to Goldschmidt's treatment of the allelomorphic sex
genes in Lymantria since 1911; it has been used since, besides in
work with sex-genes, by Wright (1925) for the series of coat
colors in guinea pigs; by Stern (19296), for the bobbed series;
by Ogura (1932), for the molting genes of the silkworm; and
by Mohr (1932), for the vgr-series.) The different types of
scalloping would then be produced by values below 40, i.e.,
insufficiencies, and these values may be read from a diagram like
Fig. 20. Using an arbitrary basis, Mohr makes the follow-
ing evaluation: No-wing = 6, vestigial = 10, notched = 15,
nicked = 22, Wild = 30 for one gene. From this the compounds
and heterozygotes may be calculated and are, according to Mohr:

vgNw vg vg vgn° vgno vgni vgno vgni + vgni + rgni + + +

VQNV, jjgSw Vg pgS W „g yg.V V VgUO „g „^.V W pgnO yg ygni vgnO vgni _|.

12 lt> 20 21 25 28 30 32 36 37 40 44 45 52 60

Wild type

This corresponds, of course, to the diagrams Fig. 20, 24 and
shows the same orderly behavior and the series of different Wild
types beyond the threshold of 40. Now, if we apply this
enumeration to the problem of dominance in the series, we see
that the combined action of a heterozygote or a compound

THE MUTATED GENE 119

is always an addition of the actions of the individual genes or,

in other words, always intermediate. The fact that Wild type

is completely dominant in certain combinations shows that the

Wild-type gene has in homozygous condition an action far

beyond the necessary threshold (see Fig. 20) ; furthermore, that

one Wild-type gene is not dominant but acts only as a dominant

if the partner in the heterozygote has an action of such a quantity

A -4- a
that — n — reaches the threshold. If, however, this partner,

as in the No-wing/ -f- heterozygote has itself such a low quantity
of action that the quantity of the heterozygote falls below the
threshold, Wild type is no longer dominant, and the lower mem-
ber of the series is called a partial dominant.

Just as the correctness of these conclusions for the right
end of the series could be proved by the parallel shifting action
of the dominigenes, so it can be proved also near the threshold
point by the action of the dominigenes upon the No- wing/ +
heterozygote, which is shifted into 100 per cent scalloping.

3. The analysis of dominance in this case can be pushed one
step further. In Fig. 20 (page 59), we presented the interpreta-
tion of the ygf-allelomorphs in terms of reaction velocities, based
upon the fact that with increase of the effect (destruction of
wing area) the onset of the process of destruction occurs earlier
and earlier in development. If we follow here only the interpreta-
tion for insufririeney of a growth substance, which might easily
be translated into the one for lysis, the topmost curve in this
figure represents the production of the growth substance just
above the threshold for Wild type. Comparing this figure with
Mohr's table on page 118, this curve would correspond to the
value 40 of the table. The curve for Wild type would be far
above this figure in the direction of the arrow, marked hyper-
irild. As we know that the curve for heterozygous No-wing
is similar to the nicked curve in the diagram, and as we have to
assume that, ceteris paribus, the curve of the heterozygote is
exactly between those of the homozygotes, the Wild-type curve
would have to be drawn as far above the m-curve as the A%-curve
is situated below it. Thus we realize that in this case dominance
is based upon a system of developmental reactions, as described
in our model (Fig. 23) ; the heterozygous gene produces an inter-
mediate reaction which leads to a phenotypically intermediate

120 PHYSIOLOGICAL GENETICS

condition, whenever the threshold for Wild type is not trans-
gressed. This threshold, however, is determined independently
by other reactions, another variable. In the present case, the
facts of development show that the threshold is identical with
the time in development at which the differentiation of the wing
An luge in the pupa is finally determined. If up to this moment,
which is controlled independently from the vg-gone, the growth
substance has been available in minimum quantity, the normal
Wild-type wing will appear. It is finally evident that domini-
genes or environmental conditions may shift one or the other
of these two known variables — as well as unknown others —
with the effect of changed dominance conditions. It is also
evident that it will be most difficult to accomplish such a shift-
ing effect if the Wild-type gene is involved in the heterozygote.
The reason is that the curve of this gene is so far above the
threshold that only in the heterozygote with the lowest, almost
lethal member of the series some individuals will remain below
the threshold.

This best studied example (next to the Bar case in Drosophila,
which belongs to another chapter, see page 143) then shows
dominance as a function of the interplay of gene-controlled
reactions and thus furnishes further evidence toward an under-
standing of gene action.

We have mentioned the different types of behavior in regard
to dominance found in series of multiple allelomorphs. The
question arises whether all these types may be interpreted on the
same basis as the type case of vestigial. Let us consider first
the cases in which each member of a series is dominant over the
lower one. It is obvious that such a case cannot be explained
with intermediate reactions and one threshold value for Wild
type. We must rather expect some type of threshold conditions
at each level. Such may easily be produced if the final reaction
is not of a plus-minus type with all grades in between but of a
strictly alternative type. Let us assume that a case of pigmenta-
tion is involved, as in certain actual cases (see Stern, 1930).
We remember from the work of Scott-Moncrieff that usually
the higher oxydized or methylated types are dominant. There is
nothing intermediate between one or two methyl residues. If,
then, two alleles differ by producing the final addition of one or
two methyl residues to the pigment molecule, one must neces-

THE Ml TA TED GENE 1 2 1

sarily be dominant. This model will account for all similar
cases, whatever the details of the chemism may be.

There is no need for a special explanation of the cases of
pleiotropy, in which each series of actions of the same gene
has its own dominance relations. We have already pointed out
(page 89) that the independent behavior of these different
effects in regard to sedation is dependent upon the special
developmental conditions of the different parts of the body.
This means for the case of dominance that the variables of our
model (Fig. 23) may be different for each organ involved with
all the consequences for dominance, already reported. The
same type of reasoning applies to those cases in which different
alleles cause very different processes, e.g., spineless aristapedia.
But here another additional problem is involved, viz., the problem
of pattern, to which we shall return in another chapter.

F. The Phylogenetic Theory of Dominance

Phylogenetic deliberations have nothing to do directly with
the physiology of gene action. But, in connection with domi-
nance, they may at least be mentioned, because the phylogenetic
argument has silently assumed that dominance is a phenomenon
of such a physiological type as had been developed in Gold-
schmidt's interpretation of dominance (1927c). As a matter of
fact, Wright (1934(2) clearly stated that this is the case and worked
out his theoretical formulation of the physiology of dominance
(see page 102) with the intention of furnishing a basis for a
phylogenetic theory.

Fisher (1928-1932) has developed the theory that the domi-
nance of Wild type over mutations is the result of a selection of
modifiers which tend to shift the phenotype of the heterozygote
toward resemblance to Wild type. He assumes that recessive
mutations must have occurred always in sufficient number and
therefore that heterozygotes were always present. If those
heterozygotes which most resembled Wild type had a selective
advantage, this would work finally to a selection of complete
dominance by selection of genes modifying this dominance.

Wright (1929) and Haldane (1930) have published objec-
tions to this theory which, after all, it will be difficult to prove or
disprove. But, as stated above, such a theory depends com-

122 PHYSIOLOGICAL GENETICS

pletely upon a knowledge of the physiology of dominance.

The former discussions make it clear that the facts that physio-
logical genetics have brought to light regarding dominance furnish
a basis for such views as Fisher holds. But the decisive problem
tor the phylogenetic side is, of course, not the underlying develop-
mental mechanism but the question of selection pressure, as
discussed in the papers of Fisher's opponents, but which do not

belong here.

Regarding the developmental system and its meaning for a
possible selection process, some additional views have been
derived, which are actually based upon the physiological facts.

Muller (1932rf) realized that a theory of the phylogenetic
evolution of dominance of the Wild type by selection of modifiers
requires a physiological interpretation of dominance of the same
type as the one developed by Goldschmidt (1927c) and proposes
to look at the selection problem from a different angle. He
postulates that the mutations of modifiers favoring dominance
have been selected not so much for their specific protection
against heterozygosis at the locus in question as to provide a
margin of stability and security, to insure the organism against
weakening genetic or environmental variability. "These modi-
fiers must so affect the reaction set in motion by the primary
gene in question as to cause this gene, when in two doses, to be
near an upper limit of its curve of effectiveness, that is in a nearly
horizontal part of the curve, not so readily subject to variation
by influences in general, including reduction in the dosage of
the primary gene." This, of course, presupposes an explanation
of dominance of the type of our model (Fig. 23) and of the type
proved to exist in the vestigial case (Fig. 20). The phylogenetic
speculations (and calculations) on the selection of dominance
modifiers then fall in line with the results derived from physiologi-
cal genetics.

Simultaneously with Muller, Plunkett (1932) derived the same
conclusion and expresses it in the same terms as those derived
above from the vestigial case. He points out that Wild type is
much less susceptible to environmental and other effects then
the mutants. This, he says, is due to differences in the distance
of the developmental process from its asymptote at the time when
it ends; processes controlled by Wild-type genes usually closely
approach their asymptotes, while those modified by mutant

THE MUTATED GENE 123

genes may be terminated by the effects of other developmental
processes while still very incomplete. Wright (1934) also comes
to the same conclusions.

But another point derived from physiological genetics has also
entered this discussion. In the vestigial case, we (Mohr and the
author) found that it must be assumed that within the Wild
phenotype many different degrees of Wildness must exist, all
appearing alike on account of the threshold conditions, a view
that had originally been derived in another field by Goldschmidt
from the analysis of intersexuality for different grades of maleness
or femaleness. In discussing Fisher's work, Haldane (1930)
points out that selection might also have selected instead of
modifiers, more potent Wild-type genes (what we called Hyper-
wild, far above the threshold). Muller (1932) set out to prove
the existence of such Wild-type genes in different potencies. He
used the findings of Timofeeff (1932) that a Russian race of
Drosophila showed a different rate of mutation to white eye
from the rate in the American race. Assuming that the Wild
allelomorphs in this case may actually be different, Muller com-
bined both of them with two white eye genes in a triploid con-
dition (attached X's). He found that the American gene was
less dominant than the Russian and attributes this to different
potency, based on different levels above the threshold. We shall
meet the same set of facts in a later chapter.

6. THE GENE IN DIFFERENT DOSES

The most elementary facts of genetics require the assumption
that it is not irrelevant whether a gene is present in one or two
doses. This quantitative view found its early expression in
Bateson's presence-absence theory which claimed that a dominant
gene is the presence, and the recessive the absence, of something.
The heterozygote therefore contained one dose of a gene as
compared with the two doses of the homozygous dominant. It
is known that this theory has been disproved, at least in this its
original form. Another type of quantitative view was intro-
duced by Goldschmidt (1917) and claims that the quantity
of the thing that is called a gene is important for the result
of genie action and that gene and effect are linked, ceteris paribus,
by the simple relation of a proportion between the quantity of
the gene and the velocity of the chain of reactions controlled

124 PHYSIOLOGICAL GENETICS

by the gene. The (frequently misrepresented) derivation of this
viewpoint was as follows. It was found that sex is determined
by a balance between male and female sex determiners, one of
them (the female one- in byniantria) being constant for both
sexes, the other (the male determiners in Lymantria) being car-
ried by the X-chromosomes and therefore being present in either
one or two quantities. These sex determiners in the X-chromo-
some behaved like single genes and were therefore considered as
such. Then a series of genie conditions was found, which
behaved genetically like a series of multiple allelomorphs of the
sex gene and which shifted the balance in an orderly seriation
from the value typical for 1 X through all intermediate values
to the value characteristic for 2X (the series of intersexes).
Since IX or 2X, i.e., one or two sex genes, obviously are two dif-
ferent quantities determining the value of the two balanced
systems F/M and F/M + M, the series of intermediate values
for the series of intersexes obtained in the experiments must
represent different quantities of the sex genes between the
quantity one (M) and the quantity two (M + M). It is cer-
tainly difficult to escape this conclusion, which is independent of
the question whether M is one or a group of linked genes. It
was strengthened by the proof that all the possible compounds
of this series of multiple alleles had exactly the balance values
that were expected from the known values of the individual
members of the series (measurable by the degree of inter-
sexuality, or sex reversal).

As it was simultaneously shown that these series of sex genes
control reactions of definite velocities (see page 52) fit to be
arranged in the same orderly way as the balance values, the con-
clusion was obvious that it is the quantity of those genes which
controls the velocity of the reactions, viz., the production of
stuffs controlling sexual differentiation.

What was up to this point a logical inference from the facts
soon was proved to be actually demonstrable. Standfuss (1907,
1908), who had produced intersexes in the saturnid moth by a
species backcross (at that time not distinguished from gynandro-
morphs), realized the correct type of interpretation. After
Federley's (1913) demonstration of triploidy in such back-
crosses, he noticed that this abnormal number of chromosomes
was responsible for the production of these intersexes. (Gold-

THE MUTATED GENE 125

schmidt and Pariser, 1923, added the cytological facts later.)
Much clearer was the demonstration in Drosophila by Bridges
(19216, 1922), where different quantities of autosomal and sex
genes may be secured, among them the ratio 3.4 :2X producing
intersexes. Hen1, then, the actual quantities of genes con-
tained in visible numbers of chromosomes were shown to account
for the abnormal development called intersexuality. Since that
time, other similar cases have been found (for details, see Gold-
schmidt, 1931d). The generalizations derived from this analysis
in regard to the nature of the gene and the multiple allelomorphs
belong to a later theoretical chapter along with the discussion
of some rather naive criticisms. Here we are interested to know
whet her other facts exist that prove that the action of the gene is,
ceteris paribus, proportional to its quantity; that an orderly
series of quantities will lead to the typical effects seen in series of
multiple allelomorphs; and that there is a relation between
gene quantities and velocities of gene-controlled reactions of the
type found in chemical reactions; and finally to learn whether
facts exist to show generally that the quantity of a gene is one of
its important properties.

A. Differences of Dosage Produced by the IX — 2X

Conditions

It is obvious that the mechanism of the sex chromosomes
permits not only" the study of sex genes in different quantities
but also the study of all sex-linked genes in one or two doses. In
the overwhelming majority of cases, one sex-linked gene in the
heterozygous sex has the same effect as two in the homozygous
sex or even a larger effect upon the phenotype. But there are
also a few cases known in which the haplo-effect effect is weaker
than the diplo-effect. Before discussing these facts and their
meaning, we want to point out another set of facts that is perti-
nent to this problem.

1. Sex-controlled Phenotype. — There are numerous examples
to show that one and the same genotype, except for the sex
chromosomes, has a very different phenotypic expression in both
sexes. We leave aside for the moment everything described as
secondary sex characters and consider only cases of mutant genes
outside the sex chromosomes, therefore present in both cases in
duplex condition. There are innumerable cases in which such

26

PHYSIOLOGICAL GENETICS

homozygous mutant genes act alike in both sexes. Bui there arc
also numerous instances in which the effect differs in the two
sexes. Some of these facts are of a type that might lead to an
understanding of the phenomenon in terms of developmental

physiology.

In Lymantria dispar, some of the geographic races are dis-
tinguished by different speed of development, different growth

2000

1800

Hichi SH

^d Hie hi

n m w v Pu

Fig. 25. — Growth curves of male Lymantria dispar, race Hichinoe, females
with five moults, males with four. Larval instars plotted against weight in milli-
grams. (From Goldschmidt, 1933, Arch. Entwicklgmech. 130, Fig. 14.)

curves, different number of molts; and all these characters are
typically different in both sexes (Goldschmidt, 1933a). But also
the degree of difference between the sexes might be different in the
respective races and might be controlled genetically. Figure 25
shows the growth curves for the two sexes of one of the races in
which females have five and males four molts; and Fig. 18 gave
similar curves for a number of races, showing the genetic differ-
ences. But, as is usually the case with growth, the difference is
not one of single genes. In regard to the number of molts, how-

THE MUTATED GENE 127

ever, the situation is relatively simple (see Goldschmidt, 1933a),
as the different types are produced by a series of three multiple
allelomorphs (parallel to the case of the silkworm, according to
Ogura). 7\ homozygous produces four molts in both sexes;
T2, four molts in the male and five in the female; and T3, five
molts in both sexes. T2 then has the same effect in the female as
T-i in the male; etc. It is further known from Ogura's work that
in the silkworm occasionally a variation into the next lower or
higher phenotype occurs, which we might have mentioned as a
threshold phenomenon: It is further known that it is possible"
to shift the number of molts to a certain extent by the action of
temperature (Ogura, etc., silkworm; Kuehn and collaborators,
flour moth) or by hunger (Goldschmidt, Lymantria). These
facts certainly reveal a genetic and developmental system
involving a series of embryonic reactions of different velocities,
capable of being shifted either by a mutant gene or by external
conditions and exhibiting the phenomenon of threshold values.
Within this system, the two sexes are found on different quanti-
tative levels; i.e., a mutant gene or an external influence may
raise or lowrer the curve of one sex to the level of the other in a
different genotype. Such systems have been described repeat-
edly, and it has been shown that the resulting phenotype is a
function of the genie action and all the other conditions of the
developmental system. The autosomal genes involved in this
case are identical for both sexes. What is different, then, is the
developmental system. The X-XX mechanism creates two
different developmental systems, probably in regard to the veloci-
ties of certain basic reactions, generally described in terms of
metabolic rate. The reactions controlled by mutant genes,
therefore, are facing a different situation in the two sexes. Let us
assume for argument's sake that one of the differences, controlled
by the X-chromosome mechanism, is the time at which differ-
entiation ends. One and the same reaction caused by a mutant
gene might then be ended at a different level in both sexes
according to the different time limit for differentiation. If this
reaction controls the incidence of a molt, it will occur or not
occur according to the sexual difference just assumed. If the
reaction consists in the oxidation of something, a higher or
lower degree of oxidation will appear in one or the other sex. If
this is true, it will follow that, the appearance or nonappearance

128 PHYSIOLOGIC \l. GENETICS

of a sexual difference in regard to the action of autosomal genes
depends upon the type of reaction involved and whether the
quantity of its products, or threshold values for its action, arc
interfered with by the different conditions of the developmental
system in both sexes.

A number of facts of the same general type as the previous
example are available, all of which tend to corroborate the
foregoing interpretation. There is a first set of facts not relating
to specific genes but to general sexual differences. We have
already mentioned the fact that it is possible to reproduce
certain different ial features of one sex as a phenocopy in the other
sex by action of temperature shocks in the critical period. As a
matter of fact, some of the earliest phenocopies produced by
Standfuss, and also later by Federley and Kosminsky, in Lepi-
doptera were of tins type (see page 4). A second series of facts
relates to the action of temperature upon the phenotype of
mutants. We reported earlier (see page 16) Driver's analysis
of the Bar eye in Drosophila and Harnly's work on the vestigial
wing. In both cases, the response was somewhat different in
both sexes. In addition, it may be stated that the sensitive
period for the action of temperature may be situated at different
times of development. This is to be expected on the basis of
such differences in the time relations of processes of differentiation
as are known to obtain for the two sexes of many animals and
which have been registered in all the work on details of growth.
A third group of facts is of special interest because it contains a
certain quantitative element. There are known mutational
changes which are caused by the simultaneous action of a
number of mutant genes with additive effects, and in some such
instances these effects differ quantitatively in both sexes.
According to Goldschmidt (1921b), melanism in the nun moth
(Lymantria monacha) is caused by two autosomal and one sex-
linked gene, which have an additive effect: if all three mutant
genes are present in homozygous state, the wings are black; in
their absence, they are white; and the different homozygous and
heterozygous combinations produce all intermediate conditions
according to definite rules. If we picture side by side a series of
males and females of the same genetic constitution with increasing
number of the three melanistic genes, we find the males always
about one class of darkening ahead of the females; i.e., a male Aa

THE MUTATED GENE 129

looks like a female A A (the mutant genes are dominant); a male
AAbb like a female AABb; etc. This applies as well to the
autosomal as to the sex-linked genes, since males with the sex-
linked gene in heterozygous condition are much darker than
females in simplex condition. Goldschmidt interpreted these
facts by saying (1920, page 95): The cause must be sought in the
physiological phenomena of pigmentation. Exact knowledge is
lacking but the facts indicate that the genes for pigmentation
control a definite increase in the quantity of pigment and that a
given quantity covers the small wing of the male faster than the
larger wing of the female. A better description would have to be
given in the terms used above.

A similar example is also available from the work on the
vestigial character of Drosophila (Goldschmidt, 1935-1937).
Here, as described on page 109, a series of three mutant genes
(dominigenes) shift the phenotype of the vg/A- heterozygote
toward scalloped wings, and again the three genes, one of which
is sex linked, have an additive effect. The effect, however,
requires the presence of both dominant autosomal genes at least
once in the female (AaBb), whereas males of the constitution
A A bb show the same effect. This case is of interest because many
of the facts concerning development, phenocopies, and multiple
alleles are known and furnish just such a system of develop-
mental processes as is required for the explanation of the phe-
nomenon with which we are concerned here.

A fourth group of facts which belong here are the facts con-
cerning sex-controlled heredity. This term (Goldschmidt,
1920a) means that the phenotypic effect of mutant genes is
suppressed in one sex, i.e., is controlled by the specific features of
sexual development. (It partly covers the older term sex
limited.) Its actual basis can best be demonstrated in certain
crosses of L. dispar. There are mutants existing in regard to the
wing color of the male, not visible in the female. In a backcross
or F2 involving such mutants, the females are all alike; the males
show the simple segregation as expected in all such cases of
crossing with sex-controlled traits. If the cross, however, is such,
a one that the females become intersexual and assume male wing
color, the segregation is also visible on the wings of the intersexual
females (Goldschmidt, 1920c). A number of instances of sex-
controlled heredity are known partly with very large phenotypic

130 PHYSIOLOGICAL GENETICS

differences, in which the underlying mutant genes have been
analyzed, viz., the case of Papilio polytes (De Meijere, 1910;
Fryer, 1913), P. dardanus (Poulton, Ford, 1936), Argynnis
paphia (Goldschmidt and Fischer, 1922), and Colias edusa
(Gerould, 1911). Here mutant genes and their recombinations
produce the phenotypes of the immensely different mimetic-
females, with normal Mendelian segregations, whereas the males
are phenotypically constant but genetically identical with the
females. In case of gynandromorphism (Goldschmidt and
Fischer, 1927), the female parts may show one or the other sex-
controlled type, whereas the male side is not influenced (see also
Cockayne, 1935).

It is obvious that these facts also require for their explanation
a developmental system, as described before in this chapter,
a system in which the time relations of determinative processes
in the development of one or the other sex are fixed in such a way
that the action of the mutant genes reaches an effective thresh-
old, or the production of the necessary amount of something,
either in time for taking part in the following developmental
processes or too late for this. In more detailed form, this idea
is developed in Goldschmidt and Minami (1933) and Gold-
schmidt (1927c).

2. Dosage through Sex Chromosomes. — Now we may turn
to the main subject of this chapter, the effect of different doses
of one and the same gene, as produced in the two sexes in the
case of sex-linked genes. As mentioned before, there is usually
no difference of effect, and this may be for a number of reasons
which are discussed below:

1. Sex-linked genes may be genes of such a high potency of
action that the maximum effect that is physiologically possible
is produced in the simplex condition. Haldane has pointed out
that such a condition in the Wild type might have been brought
about by selection in the past. But if mutant genes are involved,
or even series of multiple allelomorphs, such an interpretation is
not very probable.

2. Dissimilar effects of one and two genes in the respective
sexes may be prevented by a system of dosage compensation,
according to the idea of Muller (1932), who assumed that within
the X-chromosome a number of modifying genes exist, which
produce a compensating effect, which makes up for the lessened

THE MUTATED GEM: 131

efficiency of the gene in simplex condition. But there must
be some reason for the existence of these dosage compensators,
which cannot be expected to be present for eventual action upon
mutations. Muller therefore assumes that they have also a
similar action upon the Wild type, which thus is held high above
its minimum threshold effect, and that consequently these modi-
fiers have been evolved by selection after the manner of Fisher's
theory (but independently of heterozygosis).

3. The interpretation introduced above to explain sex-
controlled effects also explains the facts being considered here.
This seems to me the most reasonable explanation. The sex-
determining mechanism, acting through the control of the
balance of the sex genes, determines not only the course of sexual
development but simultaneously all the concomitant differences
in metabolism and rate of differentiating processes. These
differences are known to occur at the time when many mutant
genes produce their phenotypic effect. For example, in many
moths the males develop faster than the females, but pupal
development takes longer in males (exact data in Goldschmidt,
1933a). Sex-linked genes will therefore produce the same effect
in a simplex condition as in duplex, when the sex-controlled
changes in the developmental system of the simplex sex run
parallel to the change in gene action from duplex to simplex.
If, for example, the reaction produced by one portion of a mutant
gene reaches the threshold of action later than in the case of two
portions, the effect of simplex and duplex will be the same, if
the whole system of the simplex sex is such as to fit in with a
later action of the respective gene. If this is not the case,
the simplex gene will have a different effect — a lower one if the
shift is in one, an even higher one if the shift is in the other,
direction. This explanation which covers sex-linked genes both
with or without rate effect, also with increased effect of simplex,
and which fits this phenomenon into the whole system of action
of genes, as revealed in physiological genetics, seems superior
to the phylogenetic explanations of dosage compensation.

If this is the case, the difference or lack of difference in the
effect of one or two doses of sex-linked genes furnishes no direct
results for the problem of the action of genes in different quanti-
ties. The effect, whatever it is, is not produced under the condi-
tions of ceteris paribus but in two different physiological systems.

132

PHYSIOLOGIC \l. GEh ETICS

The question remains whether or not some of the detailed data
available point in the same direction. The most extensive set

of tacts known thus far relates to the eye-pigmen1 series in
Drosophila at the white locus of the X-chromosome. In Table
13, the phenotypes are arranged serially from dark to light for
t he female.

Table 13

Sym-

Simplex compared

with

Name

9

cf

duplex

bol

It-

(•dial

deep pink

still darker

increased

IP'

blood

pink

pink

equal

IP

eosin

pink

pinkish yellow

decreased

Wch

cherry

pinkish

darker

increased

IP

apricot

pinkish yellow-

darker

increased

IP

ivory

pale yellow-

lighter

decreased

IP

tinged

very pale pink-
ish

very pale pink-
ish

equal

IP/

buff

light, brownish
yellow

light brownish
yellow

equal

W

white

white

white

equal

These data show the presence of all three possibilities of simplex
effects in one series of alleles. It is remarkable that a chromosome
rearrangement exists, the so-called pale translocation, which
affects the eye colors of this series. Those that are lighter in the
male are still lighter, and those that are darker are still darker.
The pale translocation has therefore the same effect as the sexual
difference of the system of development. Other eye-color modi-
fying genes with comparable effect are known, e.g., pinkish,
which dilutes male more than female eosin (Stern, 1929a).
These facts point to the same type of explanation as before: the
effect of the modifiers certainly works through a set of reactions
different from that controlled by the white series. Here a change
in the interplay of different developmental reactions is patent.
As the effect is of the same type as the sex effect, we might
reasonably attribute it to the same causes. It must be agreed,
though, that it will be difficult to devise a detailed form of such
a system to cover also the irregularity of behavior within the
series.

THE MUTATED GENE 133

We have mentioned Ephrussi and Beadle's transplantations
of Drosophila eyes, to which we shall return later. When sepia
eye disks were implanted into vermilion hosts, a sex difference
in pigment development appeared (1937), the males being
lighter. Transplants performed with disks of different age
showed that the surroundings cannot be made responsible for
the effect, which must be caused by conditions in the disk itself.
This does not contradict the former conclusions, as the disk is
itself a developmental system in which the pigment-forming
reactions are only a part of all reactions.

B. Differences of Dosage by Deficiencies

Next to the instances of sex-linked genes, mutant genes lying
opposite deficiencies, i.e., deletions of a part of the chromosome,
provide the most illuminating material on the action of genes in
single doses as compared with double doses and the heterozygote.
Though deficiencies are known for many organisms used in
genetic experimentation, most of the information pertaining to
the present problem comes from work in Drosophila. Mohr
(1923a) discovered that most mutant genes contained in a section
of a chromosome opposite a deficiency in the sister chromosome
had in this their simplex condition an exaggerated effect; e.g.,
a light eye color was still lighter, a forked bristle extremely
forked. This applies to the majority of mutant genes thus
studied by Mohr and his followers, with the exception of mutant
characters which seem to be already at the extremest, physiologi-
cally possible level, like white eyes, prune, purple, and yellow.
A first explanation for this phenomenon was given by Bridges
(1922), who assumed that the deleted piece of the chromosome
contained a number of modifiers the removal of which upset
the balance of the genes in favor of a more extreme action of the
mutant gene. This interpretation was not very probable at the
outset, as it ought to lead in different cases to results of opposite
character, since it can hardly be supposed that the deleted pieces
always contain only one type of modifiers. Mohr, moreover,
could show that the phenomenon is independent of the length
of the deletions in one region, which also rules out the balance
interpretation. Goldschmidt (1927c) proposed a different expla-
nation, based upon the general idea that the activity of the
genes is proportional to their quantity. If, for example, the

I ; 1 /*// YS/nLOGICAL GENETICS

mutant gene for rosin exes produces a smaller quantity of pig-
iin'iit (or a lower grade of oxidation) than the wild-type allelo-
morph, the seriation of decreasing quantities produced will be
+/+ > -V /wr > w*/W > iV; i.e., one IV leads to less pig-
mentation than two w'. As the condition produced by the
homozygous recessive we/w" is regarded as the mutant type,
the effect of one 10' will be still further away from normal, i.e.,
exaggerated in comparison with the homozygous recessive.
'This interpretation, which in itself is independent of the assump-
tion that the mutation from + to we was a change in quantity
(an assumption strongly suggested by these facts), has since
been proved to be correct. The proofs, derived from the study
of the effects of more than two recessives, wdll be discussed in
the next chapter. They have been accepted as valid by Stern
(1930), Mohr (1932), and Muller (1932). The last-named
author has proposed to call mutations, which produce a lessened
effect (slow-er rate of reaction according to Goldschmidt) and
therefore show this dosage phenomenon, hypomorphs (which
seems unnecessary) and has added a number of other instances,
all of the same type. These facts, then, actually show that in
this case the effect of a gene is proportional to its quantity.

In a considerable number of the cases studied, the facts reported
above led to the conclusion that the action of the mutant genes
consists in changing- rates of reaction of the special physiological
process involved in the production of the phenotype. The
phenomena just reported then prove that, in these cases at least,
the velocities of the respective reactions are, ceteris paribus,
proportional to the quantities of what is called a gene. This
proved fact might or might not justify the extension of this view
to the primary differences between the Wild type and the mutant
gene. The author's impression is that it is not at all easy to
evade this conclusion, but the discussion of this point belongs
to the chapter on the theory of the gene. It may be added,
however, that the majority of gene mutations, according to
Muller (1932), are of this hypomorphic type.

Muller (1932) refers to a number of cases which, though
they show that the dosage of a gene has a typical proportional
effect upon the end product, yet differ in important points from
the simple relations just reported. He points to Mohr's work
on the behavior of abnormal abdomen in different doses. The

THE MUTATED GENE 135

difference from the cases just reported is that the homozygous
mutant gene has a more extreme effect than one dose plus
deficiency. The seriation here is from the standpoint of maxi-
mum abnormality: Ab/Ab > Ab > Ab/+ > +. He calls such
mutants, which apparently do not act in the same direction
as the Wild-type allele but in a different direction altogether,
though to a lesser degree, antimorphs. Leaving aside the question
of whether or not the cases ranged in this group belong actually
to gene mutations, it must be pointed out that the difference
between the two cases is not a difference of the type of mutation
but a difference concerning the processes controlled by the genes
in question, a difference in physiology of development. In the
case of the eye pigment, the mutant gene produces less pigment
or less oxidized pigment; it is deficient in action, or, more cor-
rectly, it controls a lower velocity of reaction; one dose therefore
must have the minimum pigment-producing effect. In the case
of the abnormal abdomen, the effect of the mutant gene is to
produce a situation that prevents normal segmentation or
normal concrescence of the segments, whatever the immediate
cause may be. Inspecting the different degrees of this effect,
we realize that the gradation of the effect is proportional to the
time of onset of the disturbing feature. The mutant genes then
control a reaction that leads to a threshold of action at a definite
time of development. Two doses of the gene will, just as in the
pigment case, produce a faster reaction than one dose. In this
case, however, faster reaction means earlier onset of the disturb-
ance and therefore more extreme disturbance. This deliberation
shows that the cases that Muller calls antimorphs lead to exactly
the same conclusions in regard to the effects of gene dosage as
the others, if viewed from the standpoint of genie action in
development. Further material bearing upon these problems
will be found in the next chapter.

C. Differences of Dosage by Addition of Genes

The elaborate knowledge of Drosophila genetics has made it
possible to build up individuals that contain series of different
dosages of one gene and thus to test further the influence of gene
quantity upon the product of genie action. One such case has
been available for a long time, though it was realized only recently
— the case of the Bar-eye mutants, which will be discussed shortly.

136

PR] SIOLOGICAL GENETICS

The first systematic attempt to attack the problem of the relation
of gene quantity to gene action by gene addition was made by
Stem (19296) with the bobbed allelomorphs. The Y-chromo-
some of Drosophila contains the mutant gene bobbed, the norma]

allele of which is situated in the X-cliromosomo. As hardly any
other genes are situated in the Y-chromosome, and as this has
very little effect upon development, the number of feb-genes could
be varied by adding more Y-chromosomes or Y-chromosome

d

X^fybb xb6^ibybb'

ybb'

Xbb

ybb"

Xbbxbbybb'ybb'

xW^bST

_

(a)

Xbb'xbbFybb ) 'bb
Xbb'xbb'ybb'

(b)

Xbbl

Xbblybb- X^YM^t^yiy
(C)

L.

H

Xbb

(d)

Fig. 26. — The squares on the left side: The relative quantities of the four
66-alleles 66', 66, 66", and 66' as derived from the length of bristles. X and Y
indicate the position of the alleles. The diagrams a-d: action of different combi-
nations of 66-alleles. Each combination shows the respective bristle, ab, female;
cd, male combinations. (From Stern, 1929, Biol. Ccntrbl. 49.)

fragments. In this wfay, males could be composed with from
zero to three bobbed genes, and females with from zero to two
such genes. Bobbed produces short bristles. In these combina-
tions, the length of the bristles is directly proportional to the
number of the 66-genes, up to the Wild-type length, which acts
as a threshold condition, beyond which no increase seems possible,
at least in this series. [As a modification by temperature action,
bristles of twice the length of Wild type may be produced
(Goldschmidt, 1935a)]. One might also express the result by
stating that the increase of numbers of recessive genes leads
toward and finally to the dominant Wild phenotype. The
experimental series and its result are represented graphieally

THE MUTATED GENE 137

in Fig. 26. These are incidentally the facts to which we alluded
in the last chapter as proofs of Goldschmidt's interpretation of
the exaggeration phenomenon.

A number of comparable facts have become known since.
Muller, League, and Offermann (1931) produced series of gene
quantities for such mutant genes of Drosophila as scute, eosin,
apricot by adding small fragments of the chromosome in question
containing the respective mutant locus, which had been secured
by irradiation. For most of the loci the results parallel com-
pletely those of Stern and lead therefore to the same conclusions.

It is in these same experiments on dosage changes that Muller
et al. found the less simple behavior, as reported earlier, for one-
dose relations in the case of deficiencies, in which two doses of the
mutant gene have more effect away from normal than one. The
probable explanation was given on page 135 for Mohr's example
of abnormal abdomen. Muller (1932) mentions another some-
what different case. Homozygous hairy wing is twice as hairy
as heterozygous. Adding the normal allele in a duplication does
not change the situation. On the other hand, a duplication
containing the Hairy gene added to a normal individual makes it
hairy. One might express this as dominance of Hairy wing over
one or two doses of the normal allele (Muller calls this a neo-
morphic mutation. We are speaking here, with Muller, of
genes and alleles, though probably all the cases thus far allotted
to this group are major changes in chromosome architecture, like
translocations). It is too early to consider these facts in simple
terms of dosage, but there is another case, which Muller puts
in the same category with his neomorphs and which is most
thoroughly known and therefore more suitable for an analysis.
This is the case of Bar eye in Drosophila.

Bar eye was originally described as a semidominant mutant
reducing the number of facets in the eye. Subsequently, a
number of what seemed to be mutations arose at the same locus,
partly more, partly less, reducing the number of facets. The
one that is important for our present discussion is Zeleny's
Ultrabar, reducing the number of facets to about 25 under stand-
ard conditions (Normal 780, Bar 68). The next important step
was Sturtevant's (1925) proof (see also L. V. Morgan, 1931) that
Ultrabar is Double-Bar, two Bar "genes" being located in the
same chromosome as a consequence of unequal crossing over

i;;s

/'// YSlol.oCKM. GENETICS

(L. V. Morgan). This discovery makes it possible to compare the
Bar gene in four different quantities with Normal, without any
other locus being involved. The effect of Bar is to reduce the
number of facets and may therefore be expressed in facet counts.
Table 16 is compiled from Sturtevant's data (B+ = Wild,
B = Bar, BB = Infrabar) :

Table 14
(From Sturtevant)

Number of facets

Genotype

at 25°

Number of B

Number of B+

B+/B+

779.4

0

2

B+/B

358.4

1

1

B/B

68.1

2

0

B+/BB

45.4

2

1

BIBB

36.4

3

0

BB/BB

24.1

4

0

This and other work have led Sturtevant to conclude that no
Wild-type allelomorph to B exists. The chromosome which
was known to have lost the Bar-gene by unequal crossing over
had no other influence upon Round eyes than the normal Wild-
type chromosome. The series then would give the actual effect
of one to four doses of B in terms of reduction or of production of
facet number. The reducing effect obviously increases with
dosage of the B gene. (The difference between B/B and BB has
been explained as so-called position effect, see page 304.) In an
analysis of these results, Goldschmidt (1927c) pointed out that
the results might be interpreted on the basis of reaction velocities
of a process responsible for facet formation. If we assume that
it is the number of divisions of the primary retinula-forming
cells that is decreased by the reaction controlled by the Bar gene,
we may visualize this reaction as the production of a division
promoting substance, which either reaches the necessary thresh-
old too late and therefore acts for too short a period or is not
produced in sufficient quantity to last long enough. We might
also visualize the reaction as producing something that stops these
divisions prematurely or as producing an insufficiency which
destroys already formed cells (as in the vgr-case). Embryological
facts will have to decide between such possibilities (see the work
of Margolis discussed on page 76). But in any case it is perfectly

THE MUTATED GENE

139

clear that we are facing here a series of facts showing that a
definite gene acts through controlling reaction velocities pro-
portional to its quantity. We have discussed (see page 104) the
manifold facts regarding this gene that enhance the safety of this
conclusion. We shall add one significant fact: Hersh (1934c)
calculated from Driver's data on temperature effect that the
curve of reaction controlled by the Bar gene is a sigmoid curve,
which would mean that the differences between the members
of the series in regard to this reaction do not represent merely
developmental arrests of the process at a greater or lesser distance
from a common upper asymptote but that the termination of the
process is approached asymptotically. This seems to point more
to a process of the type realized in the v^-wing than to an influence
upon cell divisions.

The Bar series has furnished additional important material
upon the problem discussed here, which is derived from the fact
that another mutation, Infrabar Bl, was found which may be
combined with all the others and may also be obtained as double
Infrabar by unequal crossing over. The facet counts of all these
combinations are available, and many have been tested at
different temperatures.

Infrabar, which in usual terminology is one of the multiple
alleles of Bar (see, however, page 143), differs from the rest of the
series in some properties. The eye facets are not only fewer than

Table 15
(From Sturtevant)

Genotype

Number of facets
25°

Number of B'

Number of B+

B+/B+

779.4

0

2

B+/B<

716.4

1

1

Bi/Bi

320.4

2

0

B+/5'5*

200.2

2

1

B'/B'B*

138 +

3

0

BW/BW

38.2

4

0

1 10

/•//) SIOLOGICAL GENETICS

in Round (Wild type but also irregularly arranged. The
strangest difference, however, is the reaction to temperature.
Whereas in the Bar and Ultrabar alleles facel aumber decreases
with rise of temperature, it increases in [nfrabar (Luce, 1931).
Special dominance features of Infrabar will soon he mentioned.

We shall consider first the effects of different quantities of
Infrabar, which have been obtained in a parallel series to Bar,
according to Sturtevant (1925) (Table 15).

We see a perfectly parallel series to the one recorded for Bar,
again showing that the Infrabar allele also acts in different dosages
proportional to its quantity.

BAR- REVERTED

NORMAL

BAR-DOUBLE

Fig. 27. — Salivary chromosomes of Drosophila at the Bar-locus. (From Bridges,

1936, Science 83.)

These two sets of facts in which the actually known different
doses of the alleles and the normal allelomorphs fit into an orderly
arrangement suggest that the Wild type, the Bar, and the
Infrabar alleles are all different quantities of what was supposed
to be a gene. The first two of these conclusions drawn by
Goldschmidt (1927c) have now been proved correct. Bridges
(1936) and also Muller et al. (1936) have shown that the Bar
mutation is an internal duplication of a small section of the
X-chromosome (Fig. 27). Two such sections then represent
Bar. and three Ultrabar or Double bar. This new fact makes it
possible to analyze the case further. A series of increasing
combinations of B+ and B represents, then, a series of two to six
doses of this chromosome section. If we assume, for reasons

THE MUTATED GENE

141

discussed above, that the action of these "alleles" consists in
destroying; facet-forming material during development (parallel
to the destruction of wing material in the vestigial case), we may
express the actual facet numbers as percentages of the number
in the Wild type and represent the action of the genes as the per-
centage destruction of facets. This is done in the first group of
Table 16, and Fig. 28 gives the curve obtained from these data
(the crosses are values of this series). From the curve we may

100
90
80

570

xB series
o Bi series
o Combination
series

3 4 S

Doses

Fig. 28. — Curve plotted for the Bar and Infrabar series. Plotting actual and
interpolated doses against percentage destruction of facets.

now calculate the eventual dosage of Infrabar. If this allele also
represents a dosage difference from the Wild-chromosome
section, it must be a dose between one and two. If we designate
as X the difference between the Bar duplication (2B+) and the
Infrabar of smaller dosage (between 1 and 2B+), the Infrabar locus
contains the dosage 2B+ — X, or one dose B+ and one B+ — X.
The different combinations of B+ and Bl contain then different
doses of B+ — X, as tabulated in the second section of the table
(second and third columns). To prove that Bl is a dosage
difference, viz., 2B+ — X, we mark the percentage loss of facets
in the B' combinations on the curve obtained for B (the points O
in the seriation of the table). The corresponding points on the
abscissa (dotted lines) tell us which dosage of B+ = 1 would have
been needed to produce this effect in the Bar series, which fur-

142

/•// YSIOLOOICAL GENETICS

dished the curve. If [nfrabar is actually a dosage difference like
Bar, a calculation of X front all these points for the Infrabar

series must give identical values. This has been done in the
second section of the table (fifth and sixth columns), e.g., for
B+/fii 3 _ IX = 2.4 A' = 0.6; i.e., Infrabar is a 40 per cent
addition to the Wild locals.

Table 16

Number

Number

add.

B+

B+ - X

B+/B+

2

B+/B

3

BIB

4

B+/BB

4

B/BB

5

BB/BB

6

B+/B<

B</B'

B+/B'B'

B'fB'B'

B'B'/B'B'

B/B>

B+/BB'

B</BB

B/B'B*

B</BB>

B/BB'

RB' BB<

B<B< BB

BB'BB'

BB/BB'

Per cent
facet
loss

0
54
|91
>94
95
97

59
74
82
95

91
93
94
95
95
95
96
97
97
97

Value

found

A

Facet

in terms

aumber

of B+

779

358

68

i:,

36

24

2.4

0

6

716

3.0

0

5

320

3.3

0

4

200

3.5

0

5

138

5

0

25

38

-0.

1

-0

5

0.

4

0

0

0

0.

2

0

0

0

73 . 5
50.5
41.8
38.3

37.8

37

27.9

26.7

26.7

24 1

The values are seen to vary between 40 and 75 per cent addi-
tion; Bl then would be about l1 2 Wild or :ii Bar dosage. This
ealeulation is made in a rough way without curve fitting, but the
result looks positive.

The next consequence of this analysis would be that the com-
bined dosages of B and B' ought also to fit into the series, i.e., give
identical values for X. The same calculations have been made

THE MUTATED GENE

143

for this group (the points (J) in the curve and the third section
of the table). A glance at the column for X shows that this is
not the case. X is mostly zero or even negative. This may mean
one of two things: Either Bl is not a dosage difference from B+
like B (in this case, the consistent results of the second section
could not be accounted for) ; or there is an additional disturbing
feature in the third section of combinations which has to be
analyzed.

We have already discussed the dominance relation of the Bar
series in another chapter, the special features of which had

17 25

Temperature in degrees C.
Fig. 29. — Semilogarithmic plot of the relation of facet number (DrosophUa
melanogaster) to temperature in two sets of homozygous females and the corre-
sponding heterozygotes. AA, Infrabar; A' A', Bar; AA', Barinfrabar. (From
Hersh, 1934, Amer. Naturalist 68.)

already induced Zeleny, Goldschmidt, et al. to special discussions.
Wright (1929) called attention to the fact that in the combina-
tions of Bar and Infrabar, with which we are here concerned, Bar
behaves as an almost complete dominant, keeping Infrabar from
expression. This is, of course, another description of the non-
quantitative results of the third section of Table 16 and of the
value 0 for X in terms of dominance. But this description is
true only for Sturtevant's experiments at 25°. Luce (1931)
showed, as already mentioned, that Infrabar has a different
temperature relation (thermophene) from Bar, the number of
facets increasing instead of decreasing with temperature. The

1 I \ PHYSIOLOGICAL GENETICS

heterozygote, however, follows closely the Bar thermophene, the
Bar gene thus being dominanl for the whole facet-temperature
relation. Figure 29 gives the curves constructed by Hersh
(1934) for these relations. It shows as one of the consequences
of the divergence of the temperature-facet curves for B{, B, and
B/B' thai at lower temperatures the heterozygotes assume more
and more the />' type and finally, just below 17°, reach it, so that
here B' appeals dominant over B. The additional fact regarding
Bi action is then its different thermophene, which leads auto-
matically to dominance of Bar at higher temperatures and makes
a simple quantitative arrangement of the data impossible, when
B and Bl arc simultaneously present. (Hersh, 1934, however,
points out that a single order for all temperatures is obtained if
not the end result — the number of facets — is used as a measure,
but the instantaneous increase in facets at all temperatures
measured by the algebraic value of the first derivative, see
page 105).

The old discussion, whether the whole Bar series is quantita-
tive or qualitative, thus resolves itself into the following: the Bar
components are all quantitative, and the series of Infrabar is also
quantitative. But Infrabar has, in addition, a different quality,
which is expressed in its different thermophene and therefore
makes the phenotype not simply dependent upon the quantities
involved, when Bar and Infrabar are both present. We know
that B is a double dose of a small but differentiated chromosome
segment. What Infrabar is, is not yet known. If it is, as the
genetic data indicate, a duplication of only a part of this region,
the lack of the rest of the region would have to account for the
different qualities. At present, the analysis may not be pushed
further. There remains, of course, the question whether or not
one may draw a conclusion from this case concerning the action
of the gene, since usually such a duplication woidd not be called a
gene mutation. This question will be answered later when
discussing the theory of the gene.

The Bar series has always served as a typical example for a
series of multiple allelomorphs and has therefore been cited too by
Goldschmidt when he tried to prove that such series are based
upon different quantities of the same gene. The proof of actual
dosage (with the necessary additions for Bl) in this case requires
a comparison with other cases in which both actual dosages and

THE MUTATED GENE 145

allelomorphic series have been studied. The case that started
such deliberations has already been discussed, viz., the sex races
of Lymantria, where the actual dosages 1 and 2 are known and
the series of alleles fits in between in an orderly way. A similar
case has been worked out by Stern (1928, 1930) in connection
with the different quantities of the Bobbed gene in Drosophila,
the effects of which were reported on page 136. In this case, a
number of alleles of the Bobbed gene (bb) in the X- and Y-chromo-
some were found, called, in Stern's terminology, X66, Ybb', Xbbl,
Xw . All of these have a different effect upon bristle length.
From the different combinations that may be built up Stern
estimated the relative activity of these alleles and found the
series

+ = 30
Ybb' = 10
Xhb = 8

ybb" _ 4

X»i = 2

It is now possible to make series of combinations of these
alleles together with dosage differences, as reported above. The
result is that dosages and alleles fit together perfectly in their
respective effects in all combinations studied. Figure 30 repre-
sents these results as a curve, in which the ordinate indicates
length of bristle (III normal, I very short) ; the abscissa gives the
different combinations of alleles and dosages arranged according
to the values calculated for them from the preceding figures.
The curve coincides with the assumption that the alleles repre-
sent dosage differences.

In discussing these facts, it has to be kept in mind, however,
that multiple alleles may be of a very different type; a trans-
location with visible effect at the same locus as a known mutation
acts as an allele, though both actions may have a completely
different base. The same applies to deficiencies. But this
again leads to the problem of the theory of the gene.

The questions put at the beginning of this chapter (page 125),
then, may be answered thus:

1. Facts are known that prove that the gene (or chromosome
segment) may act, ceteris paribus, proportionally to its quantity.

140

PHYSIOLOGICAL GENETICS

2. Such scries or orderly actions arc of the same type as the
visible effects of the majority of multiple allelomorphs.

KX+X'Y+Y*

el

u 3 i

S*x'x'r+

Y toby bby ■* y 1

a *'**

xbbxbby*

3£

"S F

x*x** , ,

Xb"XbbY bb ybb

Xbb'xbbYbb"YbbYbb'

x+x*bi „ ,

fibbfibbybb ybb

ybbybbytd

'xbblXbblYbb'c,ndXbbtXbhYbbYbb'andXbbXbbYbb"Ybb"
9 XbbXbblYbb'andXbbXbbYbb"

ybb y-bb

XbblXbblYbb'ondXbbXbbtYbb"
o XbbXbb'

xbbl\bbl

?
£<

09

•C T3 II

C e

a 2
£ §

99 •-

te ■*

2 .c

|1

a r w

«| -2 *

~ - — 5

-a to _
ou 5

43 s

o

3. A relation exists between gene quantities and velocities of
reaction.

4. The typical nature of such cases shows that the quantity or
dosage of a gene must be considered as one of its important
properties.

THE MUTATED GENE 147

D. Differences of Dosage Not Confined to Individual

Genes

Differences of dosage may also be produced by chromosomal
rearrangements, which add (or subtract) parts of a chromosome
or whole chromosomes to the chromosome set, as already stated.
In an ascending order, such dosage alterations would be:

1. Loss of a whole chromosome if viable. Example: haplo-IV
in Drosophila.

2. Loss of a considerable section of a chromosome. Example :
large deletions in Drosophila.

3. Loss of a small section of a chromosome. Example: the
typical deficiencies.

4. Addition of a small duplicated fragment. Example: the
Bar case.

5. Addition of a larger duplicated fragment. Example : many
duplications in Drosophila.

6. Addition of an extra chromosome. Example: all the
manifold trisomies in plants.

7. Additions of more than one extra chromosome up to a
whole set, i.e., subtriploids to triploids. For the dosage of one
gene the groups 4 to 7 mean always triplicates; groups 1 to 3,
simplex condition.

8. Addition of more than one single whole chromosome,
tetrasomics, etc.

9. Addition of more than one whole chromosome set, tetra-
ploidy, etc.

10. All combinations of these types are imaginable.

It is clear that in all these cases no direct conclusions upon the
action of the gene may be drawn, because differences in the dosage
of many genes are always involved. Nevertheless, in a general
way, we are dealing with the same phenomenon of dosage increase
that furnishes information upon the action of groups of genes in
different dosage. From the standpoint of developmental
physiology there is only a quantitative difference between these
types of a dosage effect. There can be no doubt, in the light of
the cases reported above, that any deviation from the typical
dosage of a gene, resulting in different reaction velocities, disturbs
the normal setup of the interplay of the different simultaneous
and consecutive determining reactions. The result is an abnor-

1 is PHYSIOLOGICAL GENETICS

mal development within t he sphere of influence of these read inns.
The more genes are involved in such dosage differences the larger

this sphere of influence will be, and the more general the effect.
This conclusion, of course, is a direct consequence of the idea that
genes act together by controlling rates of reactions in develop-
ment, which have to be in proper tune in order to produce the
typical effect. The working of such a system to control develop-
ment was first conceived in a more generalized way (not related
to gene action) by Guyer (1911). It was later derived from
actual experiments in change of gene dosage on the basis of
Goldschmidt's discovery and analysis of the phenomenon for
which he coined the term inter sexuality (1912, 1915) and was later
elaborated as a general principle of physiological genetics (19206).
Bridges (1922) coined the term genie balance, which is frequently
used to describe these facts, though he did not think of minutely
interwoven reaction velocities but of an effect of different genes
pulling the phenotype in a plus or minus direction. This seems
too crude a way of describing the case. In using the term genie
balance, it must certainly be kept in mind that the balance, or, in
Goldschmidt's terminology, the quantitative relation or the tune,
is not one of the genes but of the gene-controlled reactions.

In this way, the dosage relations mentioned above are supposed
to lead to a considerable upset of the balance of the gene-con-
trolled reactions concerned with development. It is in this
sense that facts of that type will lead to insight into the physiology
of genie action. It is not possible to recount all the phenotypic
effects caused by those compound changes in dosage. But there
are a number of facts that might furnish some information
toward our problems.

There is first one set of facts in which the action of individual
genes is studied but where the quantitative steps have been
produced by addition of whole chromosomes or chromosome
sets. Here are a few examples to which ought to be added
innumerable cases in which the same mutant gene has been
compared in haploids, diploids, and tetraploids.

Mangelsdorf and Fraps (1931) studied the vitamin content of
corn, which is higher in yellow than in white corn. As the
endosperm is known to be triploid, it is possible to build up
kernels containing the recessive genes for white in 0,1,2,3 doses
in the formulae (F = yellow, y = white) yyy, yyY, yYY, YYY.

THE MUTATED GENE 149

The phenotype of these classes is white, pale yellow, dilute
yellow, deep yellow. An assay of the vitamin A content of these
four genotypes actually showed its quantity directly proportional
to the number of 7-genes, viz.:

Table 17

mber of F-genes

Vitamin A Units per Gram

0

0.05

1

2.25

2

5.00

3

7.50

Each F-gene, then, produces about 2.5 units of vitamin.

Similar examples may also be taken from other endosperm
work. Thus, Lindstrom and Gerhardt (1927) found a cumulative
effect of the number of starchy genes in maize. As a matter of
fact, such relations were already indicated in the first work done
on endosperm characters by Correns (1899). A special study
was recently made by Rhoades (1936). The case is somewhat
complicated in detail. The main feature is that dotted aleurone
occurs on corn kernels when a gene Dt is present together with the
gene a for colourless aleurone and three more primary genes.
The decisive point is that kernels on the same ear may be com-
pared which are identical in all the other genes but contain
1,2,3 a-genes. The result in regard to the number of spots was a
simple linear progression with the number of ai-genes (different
numbers of the Dt-gene were, however, not additive). Another
example is Nawashin's (1930) work on allotriploid hybrids of
Crepis tectorum and C. alpina. The former species has short
achenes; the latter, long ones. In the allotriploids, as compared
with the parents and Fx, a definite seriation of the length of the
achenes is found, strictly proportional to the number of alpina
(a) and tectorum (t) chromosomes, viz.,

It < 2t + la < t + a < It + 2a < 2a.

Gowen (1933a) found additive effects similar to those reported
in corn when comparing the effects of the Hairy gene in Drosophila
in diploids and triploids. Bridges (1922) and Schultz (1935)
worked on the fourth chromosome of Drosophila. The latter
used the fourth-chromosome character shaven which shortens
and removes bristles. He compared the effects in different
dosages in diploids and triploids, and the results are proportional

i:><>

I'UYSlOUxncM. CENETICS

to the quantity considering the diploidy or triploidy of the other
chromosomes, as Table 15 shows:

Table is

(From Schultz)

2A

SN

Haplo IV

Ebctreme sv
Dies

Diplo IV

sv/ +

Triplo IV

sv/sv/sv

Type sv
Ebctreme sv

+
Slighl sti

Tetra IV

sv/ sv/sv/sv

Extra bristles
Slighl so

Dosage differences of this type will be found in many cases of
polyploidy and hyperploidy, and examples may be cited from
innumerable cases in plants, in which individual genes are com-
pared in different dosages within these systems and produce
quantitatively increasing effects. Some important facts derived
from the study of tetraploids by Lawrence and Scott-Moncrieff
will be reported in a later chapter.

At this juncture, it ought to be pointed out that in all cases
where multiple factors show a simple additive behavior a simple
quantitative relation may be involved. If we take, for example,
Nilsson-Ehles' classic case of the color of oats, the quantitative
effect of one to six multiple genes may be based upon each
gene's contributing something to the amount of oxidase produced
(or chromogen or time of onset, etc.). In most cases, a further
analysis is not possible. Where, however, the individual genes
have been isolated, as in the nun-moth case, (see page 128) and
their individual action is known, conclusions upon the action of
different quantities of similar genes may be drawn. There are,
of course, many pitfalls to be reckoned with. To mention only
one: Rasmusson (1933) finds that multiple genes do not act in an
additive way but more in the form of an interaction with a
combined effect lower than required if additive. Such a result
may mean very different things in terms of gene action, such as a
threshold or saturation effect, as in the bobbed case of Stern; a
mutual dependency of different reactions leading to the same
end product; saturation effects only within individual members
of the series; disturbance of time relations; disturbance of
reactions by others. In any case, conclusions must be carefully
weighed.

THE MUTATED GENE 151

There follow now the cases in which one chromosome may be
obtained in simplex, duplex, and triplex condition, without
regard to individual mutant genes. Such is the case, for exam-
ple, for the fourth chromosome of Drosophila (Bridges). As
this is small and supposed to contain few genes, the effects ought,
to be of a comparatively good dosage type. A haplo-IV fly is
distinguished from a normal one by smaller body size; shorter
and more slender bristles; paler body color; darker thoracic
pattern; larger and rougher o\(^; slightly spread, blunt, and
cloudy wings. A triplo-IV fly, containing three fourth-chromo-
somes, shows small and smooth eyes; narrow, more pointed wings;
darker body color; suppressed trident pattern. In other words,
the three dosage types 1,2,3 fourth-chromosomes form an orderly
series of quantitative expression of a number of phenotypic
traits in the order subnormal-normal-hypernormal expression.
It may be added here that the presence of the mutant gene
eyeless in this chromosome leads to a comparable dosage effect
for this gene alone.

Other examples of the same type may be found in Blakeslee's
(since 1921) work on Datura. The interpretation in terms of
velocities of gene-controlled reactions is the same as given above
for the action of individual genes. The difference is that there
only one or a few interwoven reactions were changed in their
time relations, and correspondingly some special somatic traits
were shifted proportionally to dosage and affected reaction
velocity. In the case of whole chromosomes, which are' probably
concerned with many, and therefore decisive, reactions, general
developmental features are affected, and therefore abnormalities
of practically the whole animal will result. In plants, the
phenotypic expression of such a dosage change will appear
predominantly as a change of the whole habitus which becomes
visible at once and which might be described in quantitative
terms for each discernible character. It is important to mention
the fact, first discovered by Blakeslee in Datura and later found
in other plants, that the phenotypic effect of a triplex condition
of one chromosome is typical and different for each chromosome,
showing that the genes affecting general developmental processes
like size, type of branching, and growth type are not distributed
at random among the chromosomes.

152 PHYSIOLOGICAL GENETICS

The genic-balance-theory, in the form used by Bridges,
expresses these effects as follows. The plus and minus modifiers,
the balance of which determines for each character its condition
in the Wild type, are distributed at random along the chromo-
somes. It' one chromosome is removed, it will usually happen
that more plus than minus modifiers (or vice versa) for one or for
several characters will be removed. Consequently, more of the
other type will be lefl in action, and the character will be shifted
in that direction. Apparently, not much meaning can be
attached to this form of description as far as the action of genes
in controlling developmental processes is concerned. It seems
advisable, therefore, to discard the term genie balance in favor of
balance of gene-controlled reaction velocities.

A special problem of dosage effect is presented when whole
chromosome sets are involved in polyploidy, and therefore the
dosage of all genes is increased concomitantly. What may be
changed in this case is (1) the proper relation of gene action (if
proportional to dosage) to the cytoplasmic substratum. (2)
The threshold for the maximum effect of gene doses may be
different for different genes; therefore at a higher dosage some
may increase their effect proportionally, others not (when their
threshold is reached). This, again, would result in a disharmony
in the time of the reactions. The effects of polyploidy might
therefore also be used to a certain extent in analyzing genie action.
It is generally known that polyploids might be more or less differ-
ent from the diploids. Muentzing (1936) has recently reviewed
all the known cases and compiled an extensive table showing the
effects of abnormal number of chromosomes in natural chromo-
some races of plants. He states that hardly any cases are known
in which no differential features from the normal diploid could be
found. He t hen compares the natural polyploids with the experi-
mental ones and comes to the conclusion that the effects are the
same. It is generally known that these effects are mostly
quantitative in regard to all parts of the plant, producing the
gigas type, though the details may vary in different cases. It
is generally assumed that these differences are the result of the
increase in cell size as a consequence of the existence of the
karyoplasmic ratio, making cell size proportional to chromosome
number. But there are also exceptions due to the fact that there
is an optimum beyond which chromosome number may not

THE MUTATED GENE 153

increase without impairing the vigor of the plant. The position

of this optimum differs in different cases. It would be difficult
to draw direct conclusions upon genie action from such fads.

There are, in addition, numerous physiological effects of
polyploidy, which, in part at least, may be directly caused by the
change in cell size. But it is questionable whether all effects
may be caused in such a way. From the long list of facts that
Muentzing has compiled from botanical literature it follows that
most quantitative traits are influenced; e.g., the growth curve
and the time of flowering are retarded. According to Wettstein
and his students, this is a consequence of a change in rate of cell
divisions. In other cases, the osmotic conditions are changed;
the starch or vitamin content is increased; resistance to cold is
heightened; general metabolism and chemical composition are
changed. The few cases in animals (Artemia salina, according
to Artom, 1920, 1926; Gross, 1932; psychid moths, Seiler, 1927)
seem to show parallel traits. Most of the authors agree that all
these effects are in the end caused by the change of cell size. If
this is the case, the physiological facts do not lead further than
the morphological ones, in so far as our problem is concerned.

But there is a more indirect way in which the study of poly-
ploids has contributed to the knowledge of the action of the gene.
In his extensive experiments on polyploidy in mosses, Wettstein
(1924b) found that in pure races the experimentally produced
polyploids show a simple geometric increase of cell volume writh
chromosome number up to the possible limit of about 4n. This
limit, in regard to addition of more chromosome sets, may,
however, be shifted considerably if hybrids are used for the
experiments. In this case, up to 16 chromosome sets could be
attained. This is possible because in such hybrids the increase
in cell size is regulated so that the additional growth is no longer
proportional to the chromosome number but decreases with
further addition of genoms. This regulation of cell size occurs
according to a definite curve (Fig. 31) which reaches its maximum
asymptotically. The equation of this curve is identical with the
equation of a monomolecular reaction. Wettstein had formerly
shown that genes exist in the pure species that control the basic
cell volume as well as the proportionality factor for its geometrical
increase with chromosome number and that these genes differ
in both species involved (Funaria, Physcomit rella) . He assumes,

154

PHYSIOLOGICAL GENETICS

therefore, that in the hybrid these two genes work against
each other and thus control the asymptotic increase in cell
volume. A- the curve of the combined effect is the curve of a
monomolecular reaction, he concludes that the action of the
genes in this case is of tins general type and follows the mass law

/*- —

Uni Bi Tri Quadrl Octo Sedecivalens n— *•

Fig. 31. — Curves for cell volume in polyploid mosses (n to 16n chromosomes)

.... and pure species, — -- hybrids with regulation. (From Wettstein,

1924, Biol. Zentralbl. 44.)

of reaction. This conclusion then coincides with those derived
from many other facts.

E. The Heterozygote as Dosage Difference

In a former chapter, the facts relating to the gene in hetero-
zygous condition wen' analyzed. We mentioned there the
presence-absence theory and also the possibility that a pair of
genes differ quantitatively. In cases of multiple alleles, it could
be said at least that the activity or potency of a pair of alleles
was different in quantity, whether this meant quantity of the
genes themselves or not. It was then found that in cases of
actual quantities, as in the Bar case, the heterozygotes fitted
into the seriation. All this suggests that mutant differences are

THE MUTATED GENE 155

frequently dosage differences between the two alleles. Facts
are therefore of interest that show that the action of the hetero-
zygote shows simple quantitative relations to that of the homo-
zygote. Many such cases have already been mentioned in the
chapters on rate genes and on multiple allelomorphs. One, in
which the type of reaction has been analyzed, might be added
here, viz., the pigmentation in the Himalaya rabbit. Here only
the "akra" of the body (tips of ear, nose, paws, tail) are pig-
mented according to a multiple allelomorphic series of colors.
W. Schultz (1915) proved that after plucking hair on the white
parts of the body, pigmented hair of the expected color is regener-
ated, if the temperature is decreased below 11°C. Iljin (1926)
found this temperature different for the different akra, and it may
be modified by experimental changes of skin temperature
(Schultz, 1929; Iljin, 1932) produced by cutting the sympathicus
or by hyperthyroidism. -Even different races with different
critical temperature were found. Engelsmeier (1935) showed
that these differences are based upon differences in skin tem-
perature and that the critical skin temperature above which no
pigment is formed is constant though dependent upon the
genotype, e.g., an/an, 33.2°C; an/a, 30.6°C. This would mean
that the Himalayan gene produces a pigment-forming reaction
with a phenotypic alternative according to the skin temperature.
As the akra are normally the coldest points, this type of mosaic
pattern is the result of the alternative reaction norm together
with the outside conditions (cooling of akra).

Schultz (1928) succeeded also in producing the blackening
effect in explants of skin. In this case, it is not the temperature
at the time of pigment formation that is decisive but the tem-
perature at the time of exposure before actual pigment formation.
It might therefore be assumed (Engelsmeier) that the oxidase is
formed only at low temperature and that its amount is dependent
upon the time of cooling as well as — with time constant — upon
the genes present (one or two quantities). Actually, with the
formula an/an, 5 min. in 25° sufficed for pigment formation in
explanations of white skin, whereas an/a individuals needed
10 to 20 min. cooling. These experiments demonstrate as well
the alternative norm of reaction as its quantitative side as
correlated with the quantity of a process (probably enzyme
production) and its dependence upon the gene quantities. There

L56 PHYSIOLOGICAL GENETICS

is an additional difference in the reactivity of back and belly
skin.

It is difficult to decide whet her or not we may actually speak
here of dosage differences. This is, indeed, permitted only if the
normal allele is ruled out. Ordinarily, this would not be per-
missible. But it is quite possible that in some cases the effect
of one +- gene upon a certain character is below the threshold
and therefore equal to zero in terms of phenotypic effect. The
presence of one or two doses of the mutant gene would in such
cases actually amount to a dosage difference. It will be difficult
to decide when such a situation may be assumed. A criterion
may be found in the actual proof of different rates of reaction
in the heterozygote and mutant homozygote.

7. THE INTERACTION OF THE GENES

One of the first facts that were realized in the first decades of
Mendelian work was that the phenotypic expression of an
individual mutant gene depends upon the condition of all the
other genes that might influence the character. A mutation in
one of these other genes will collaborate in controlling the
phenotype. This insight, which was derived partly from the
earliest classic work on plants, partly from that on rodents, found
a different expression in general terms. It formed the basis of
such representations of genie effects as the ones used first in
Goldschmidt's textbook (1911), viz., dichotomic determination
tables. It was expressed in the earlier textbooks of genetics
(Johannsen, Baur, Goldschmidt) by saying that in describing the
effect of the pair A a, tacitly BBCCDD, etc., ought to be added.
The later development of genetics has not changed this general
position though a different terminology became established. The
sum total of other genes influencing the phenotype is frequently
called the internal genetic environment (Tschetverikow) ; or if a
main mutant gene may be discerned controlling a definite char-
acter, the others, which all together form this genetic environ-
ment, are called modifiers. Thus, of course, each mutant gene
may act as a modifier for any other, though such actions will not
always become visible. But the more thoroughly the effects of
individual mutant genes become known the more the influence of
the other genes becomes visible and may be expressed generally
in terms of modifiers, if these modifying genes cannot be described

THE MUTATED GENE 157

by a visible action of their own. Many examples of this type
have been mentioned in the preceding pages. We have discussed
also the development of those genetic formulations into formula-
tions in terms of genie action, the theory of the balance or proper
tune of gene-controlled reactions.

In this field, there is a considerable body of special facts which
may be significant in connection with the problems of this book.
Certainly not all cases of genie interaction can be discussed, as
this would amount to a review of all the details of special genetics.
( )nly certain facts will therefore be mentioned that throw direct
light upon the problem of genie action or that have been found in
connection with studies devoted to this problem.

As mentioned before, the colors of the hair of rodents represent
one of the characters that has been used for a study of genie
interaction since early Mendelian days. A considerable number
of facts have already been reported in other chapters (see page
88). Referring to these, we may now consider the additional
data that apply to the problem of genie interaction. Wright has
emphasized this side of the problem, and his analysis (1927)
is the most complete one.

Wright's studies on rodent colors have been mentioned in
connection with the problems of multiple allelism (page 87) and
also in regard to the chemism of genie action (page 91). In
the same work, also, the interaction of different genes controlling
the skin color was studied. The main factors influencing
pigment belong to a black (sepia) series and to a yellow series.
There is, in addition, the albino series of allelomorphs, which has
been discussed; a dilution series; and others that control the
pattern of pigmentation. Altogether seven series of allelo-
morphic factors and many of their combinations are known.
The combinations of the albino series with sepia and yellow,
respectively, have already been reported (page 88), and Wright's
interpretation has been given (Wright, 1925). These same series
were combined with a dilutor, the pink eye factor p, the brown
factor b, and the dilution factor /; and the color grades were
measured with the color wheel. Among these effects the first
to be considered is the difference between the yellow and the
black series which is caused probably by the production of two
melanins. This production is, however, not completely inde-
pendent. Factors like b and p may act on yellow or sepia

158 PHYSIOLOGICAL GENETICS

production alone. The genes e and a decide between the produc-
tion of one or the other. Others like c and s may act upon both
through an antecedent process. The albino series determines
differences in quantity of pigment irrespective of quality and is
therefore put in the last group. It has been mentioned that the
order of effects is not the same if the albino series is combined
with sepia or with yellow, and, moreover, the seriation of effects
is not identical for colors of eyes and hair. In Wright's own
words:

Factor C is a condition for the most intense colors of both series
(red and black), factors Ch and Cd determine very nearly the same grade
of dilute yellow but widely different intensities of sepia. With factors
CT and C° no yellow whatever develops but with Cr there is an even more
intense sepia than with Cd in the fur (less intense in the eyes) while
with O there is no visible sepia at birth though a small amount develops
later in the skin and fur under the influence of cold.

Table 19 contains a partial summary of these data in terms of
color-grade evaluations, only the homozygotes (no heterozygotes
and compounds) being considered.

The explanation that has been proposed was that the albino
series produces a graded series of effects on a single process
essential to all pigmentation; furthermore, that the irregularities
of the different series are explained by the assumption of a
different threshold of effectiveness of the immediate products
in the case of yellow or sepia; since each of these is determined
independently, each individual gene behaves in this respect as an
independent unit. These thresholds may also be different
regionally (eye and fur). In addition, the assumption is made
that there is a competition between the yellow and black processes
in the black forming parts of the fur. The gene b modifies the
pigmentation of the sepia series in such a way as not to affect the
thresholds of sepia and yellow or the competition between them.
P, however, has a different effect in regard to the order in the
series, from which it is concluded that it raises the threshold of
sepia from a point close to the gene level of Aa nearly to Cr
(besides weakening the pigment-producing powers — dilution —
of the substance involved). The factor / reduces the intensity
of yellow and causes complete absence of the kind of sepia
determined by p. Both effects are supposed to occur hide-

THE MUTATED GENE

159

pendently of those already mentioned. These results may be
summarized in graphic representation in Fig. 32 (from Wright).
The average grades of sepia (solid), brown (dot and dash), and
yellow (dash) are given in the combinations of the albino series

Table 19

(From Wright)

Average grades of sepia and brown in different combinations

C series

FPB

sepia

FPbb
brown

FppB
pale
sepia

Fppbb

pale

brown

ffPB
sepia

ffPbb
brown

C

21

20.10
16.88
20.09
0

15.63
14.61
14.18
15.47
0

9.7

9.08

4.88

2.58

0

8.29
7.22
5.43
2.82
0

20.91

19.0

16.79

0

16.0

CkCk

CdCd

15.0

CrCT

CaCa

0

Average grades of yellow in different combinations

C series

FF
yellow

Ff
yellow

ff

yellow

C

10.57
7.10
6.97
0
0

9.92

6.46

0

0

6.77

CkCk

1.20

CCd

1.60

C'Cr

0

cc°

0

Average grades of pale sepia or of yellow

C series

FFppBB
pale sepia

FfppBB

pale yellow

sepia

ffppBB
cream

C...
CdCd.
CrCr.

9.70
4.88
2.58

7.60
3.40
0

2.15

0

0

and the/, b, p gene mutations. E means black; e, yellow parts
of the fur. This analysis of genie interaction, which thus far is
the most elaborate one, again shows the genes at work in the way
expressed before as a system of reaction velocities in tune.

As a kind of supplement to this analysis it might be mentioned
that Hadjidimitroff (1933) found that the amount of pigment

160

I'll YSIOLOCK \\L CENETICR

deposited in a rabbit's hair is constanl for a given genotype.
Therefore, the phenotypic expression is correlated to the length
of the hair. Longer hair showing more diluted color.

,cc

ks-.lt

CKC

df^d

cac

CrC'

cacl

Fig. 32. — Action of color factors in the guinea pig. The average grades of
sepia (solid) brown (dot and dash) yellow (dash) found with particular combina-
tions of factors of the series Fj, P, and B with the five homozygotes of the albino
series. Factor E is used to indicate reference to black parts of the fur, and e to
yellow parts of the fur. (From Wright, 1927, Genet. 12, Fig. 8.)

In connection with this work, another type of genie interaction
which has been studied in rodents may be mentioned, and the
related facts might also find their place in the chapter on domi-
nance. One Mendelian character in rodents upon which much
work has been done since the first studies by Cu£not is Spotting
in its different aspects. Spotting is usually caused by a major

THE MUTATED GENE 161

gene which prevents pigmentation in certain areas (see Onslow,
page 91). There is a considerable influence of nongenetic
factors on the variability of the character, studied especially by
Wright et al. in guinea pigs. There is further the fact that
pigmentation in these cases is dependent upon the temperature
and may be changed considerably, especially in the so-called
Dutch pattern of pigmentation of the tips (ear, nose, feet) by
temperature action. Actually, exposure to cold leads to pig-
mentation of otherwise unpigmented parts (W. Schultz, see
page 155). There is a considerable amount of genetic modifica-
tion of the pattern, i.e., the extent of the white areas. Modifying
genes may be accumulated by selection and result in almost
self-colored or almost white condition. (Most of the literature is
found in the papers of ( astle, Wright, Dunn). In a recent paper,
Dunn and Charles (1937) showed that, in mice at least, other
genes for spotting exist, which cannot be called modifiers, because
in the absence of the main gene they also cause spotting. These
give a combination effect with the main gene, and the cumulative
action of the different spotting genes determines the phenotype.
This, then, is a case of multiple-factor action, known in innumer-
able cases of quantitative characters and mentioned above also in
connection with pattern types in the case of the nun moth (page
128). In the mouse case, this cumulative action may appear
phenotypically as a change of dominance of the main gene: S/s
(s is the recessive spotting gene) is Wild type; in the presence of
another spotting gene (of the iC-group), spotting is somewhat
dominant. Of course, this is not a shift of dominance but an
additive effect visible also in the heterozygote.

These facts lead to the following deliberations. The inter-
action of different genes in controlling the phenotype may
oscillate between two extremes.

1. The genes in question exercise their main effects upon
different reactions which do not interfere with each other. In
this case, the phenotype will show a combination of the different
traits involved. Such cases have been worked out quantitatively
by Csik (1934a) for wing characters in Drosophila, many of which
may be independently combined without interfering with each
other. Such an effect may occur on the basis of qualitatively
different reactions or in connection with actions at different times.
But it is probable that such cases will prove to be rather rare (in

162 PHYSIOLOGICAL GENETICS

spite of the immense literature upon simple recombinations of
genie effects) if thoroughly investigated, because it is highly
probable that somehow a change in developmental reactions
caused by a mutant gene will create a situation that affects also
the workings of simultaneous or consecutive processes.

2. All transitions will be found from such a situation to the
other extreme, viz., that the reactions caused by different genes
act directly or indirectly upon the same process. An indirect
action would occur, for example, if one gene-controlled reaction
changed the pH value present when the other gene-controlled
reaction occurred. A direct action might be of various types as
follows:

a. One reaction produces a substance necessary for some
morphogenetic process; the other, another substance with the
same type of action, e.g., two different types of growth hormone.

b. Both reactions affect different developmental processes,
involved in the formation of the phenotype; e.g., one increases the
rate of cell division; the other, cell growth; and their effect upon
size is therefore additive.

c. Both reactions have an identical effect, reached possibly by
different means; e.g., a definite substance is as well produced by
oxidation of something as by reduction of something else. The
result again would be additive.

d. The different genie reactions affect different phases of
general developmental processes; e.g., one changes growth in
length; the other, growth in width; the result might be additive,
but it might also be a compromise between both effects.

Other possibilities may easily be visualized. In a general way,
we may therefore expect interactions to be either combinatory or
additional or of the nature of a compromise or of a combination
of these three possibilities. The actual effect will depend upon
all those variables of gene-controlled reactions that have already
been described, viz., velocity of the main reaction, threshold
values, time of final determination, time of onset of different
decisive reactions. Examples for such interactions may therefore
be found at many places in former chapters.

A systematic attack upon these problems has been made also in
Drosophila. Schultz (1929) used the so-called Minute dominant
mutations. (As a matter of fact, many if not all of these are
actually deficiencies; for the present discussion, the nature of the

THE MVTATED GENE 163

gene in question is irrelevant.) The}' arc lethal when homo-
zygous; and all of them, scattered over all chromosomes, have a
similar phenotypic effect, viz., shorter bristles, blunter wings,
plexus venation, smaller size, slow development, larger and rough
eyes, poor viability. Schultz argued that a combination of two
Minutes ought to give a lethal effect or at least an extreme
phenotype if all are concerned with the same primary reaction,
i.e., ought to give an additive effect. No combination gave such
a result. To test whether an all-or-none or threshold effect
was involved Schultz made experiments that led to the following
observations :

1. A duplication suppressing a known Minute effect at the same
section of the chromosome does not influence another Minute
effect at a different locus.

2. One Minute is recessive to two normals in triploids. Two
simultaneous but different Minutes do not change the situation.

3. Two of the same Minutes in a triploid, however, are lethal.
From such facts Schultz concluded that no threshold effect can

be involved and the different Minutes do not affect the same
primary reaction. He then combined different Minutes with
different allelomorphs of Delta, which has a phenotypic effect
that is roughly the contrary of Minute. A combination Delta-
Minute is either an extremely abnormal individual or lethal.
This effect varies quantitatively in proportion to the effect of
the Minute involved. And it varies also in proportion to the
grade of the Delta allele; the highest alleles produce, in this
combination, lethality; the next lower one, early death of the
thoroughly abnormal fly; the next lower one, later death; and
the lowest member, an abnormal fly not capable of hatching.
Similar effects were produced in combination with other genes,
showing that the effect of the Minutes concerned in these results
(the secondary reaction of Schultz) is quantitatively different in a
definite seriation of Minutes. This applies to measurable
characters like viability and bristle length within the series of
Minutes and the parallel series of combinations with other
genes. Schultz accounts for these facts by the assumption that a
reaction exists in development, in which a great many genes are
involved; and that changes in any of these contributors may set
off a secondary reaction, the Minute reaction. This would then
illustrate one of the possibilities of genie interactions, as outlined

164 PHYSIOLOGICAL GENETICS

above. But it ought to !><• added thai the destructive effect
upon the development of practically the whole organism sets
aside sucli dominant mutations from what is ordinarily called a
gene mutation. It will he wise at present to be cautious aboul
drawing conclusions from such cases as Minutes, Deltas, etc.

Dunn and Coyne (1935, 1937) have carried the same problem
•a little further. rl ney selected Minutes that show an increasing
series of retarding effects upon development, the descending
series being Mw — Mb — Mz — Mh — Ml2, which meant an average
delay in the Larval period from 3 to 1 days. These Minutes were
combined with mutant genes reducing the eye size, under the
assumption that a decrease of eye size means a change in the rela-
tive growth rate between eye and body. In these combinations
with the Lobe mutant, the eye size was actually reduced, and in
addition the grade of the reduction paralleled the foregoing series
for the growth-retarding effect of the Minutes. Apparently, in
this case the interaction effect of Minute and Lobe is clear: it is
the growth-retarding influence of the Minute reaction that
interferes with eye development of Lobe, pointing to a similar
effect of both "genes," which becomes additive.

A corresponding piece of work has been accomplished with
the eye-color mutants of Drosophila, though no actual quantita-
tive work has been published thus far. The many genes for eye
color situated at different loci have been bred in many combina-
tions showing their interaction. In a general way, it may be
said that combinations of two colors are at least as light as the
lighter one and frequently lighter. Vermilion and garnet, for
example, give together a yellowish red. Other combinations of
two colors appear white. (The same has also been found by
Whiting, 1934, in Habrobracon.) Attempts at further analysis
in terms of genie products have been made by Wright (1932),
Crew and Lamy (1932), Schultz (1935), and Ephrussi and
Beadle (1937). Wright noticed that the double recessive of
scarlet and brown, neither of which changes the eye color very
considerably, is white. Scarlet and vermilion produce together
a color like vermilion. From which it might be expected that
vermilion and brown together would produce white, which is the
case. Wright thinks that the vermilion and scarlet genes are
concerned with two different links in the same reaction chain,
whereas brown may have to do with a qualitatively different

THE MUTATED GENE 165

process. Crew and Lamy (1932) found a similar case in Droso-
sophila obscura. There is an autosomal recessive purple which
acts upon vermilion as a modifier of dominance. The double
recessives are again white, as in Wright's case and as in the
apricot-ruby combination mentioned by Agol (1931). Crew
and Lamy think that the result is to be explained on the basis of
different times of action of the two genes: Vermilion causes
premature cessation of pigment production; purple, a delay in
starting the processes, which, combined, means no pigment
production. We have reported (page 33) on the work of
Schultz who found that the eye colors fall into distinct groups
regarding the earlier or later onset of pigmentation. Brown
belongs to the group of fast formation of pigment; vermilion and
scarlet, to the slow group. In the combinations that Schultz
made (the complete data have not yet been published), he
expected that members of the same group which are supposed to
affect the same reaction should in combination not show much
difference from the effect when single. Members of different
groups, which are concerned with relatively independent reac-
tions, when combined should show marked differences from
either individual effect. In development, deposition of pigment
should begin at the time characteristic for the earlier component.
The type of pigment should be determined by the later formed
pigment. As an example, Schultz mentions the combination
vermilion-sepia. It has much pigment in the primary cells, little
below the basal membrane, like vermilion. The color of the
pigment is yellow, as in sepia. Schultz reports that 100 such
effects have been studied and found to conform to the rule, on the
basis of two independent reactions involved. The interactions
are summation products of independent reactions: each gene
does its own job. But obviously this cannot be the whole story,
as the preceding case of vermilion-brown = white indicates.

On page 186, we shall describe the results obtained by Ephrussi
and Beadle with eye transplantation in Drosophila, demonstrat-
ing in some cases the existence of substances acting at a distance.
These investigators postulated two such substances (with a recent
addition of a third one) involved in Wild-type color, viz., cn+-sub-
stance, which is missing in a cinnabar fly; and a ^-substance, a
vermilion fly being deficient in both. Vermilion and cinnabar
disks in Wild type, therefore, develop Wild-type color, whereas

166 PHYSIOLOGICAL GENETICS

other eye-color types develop their own color after transplants
tion. But vermilion and cinnabar disks grown in claret, carna-
tion, and a IVw other mutants develop either vermilion or a color
between this and Wild. The authors (1937c) then proceeded to
test eyes containing two recessive colors, one being vermilion.
It has been mentioned previously that vermilion-garnet is
yellowish. This would mean that there is a considerable
insufficiency of /'' -substance, which could be tested in appropriate
transplantations. The results were:

1. live disks of double recessives, one being vermilion, im-
planted into Wild-type hosts. In all cases, the color was that
of the other mutant (not vermilion).

2. Eye disks of the second mutant implanted in vermilion
hosts. Some develop their genetic color; others, a lighter color;
and some, a much lighter color; but none so light as the double
recessive. (There is also a sex difference.)

3. P]ye disks of the different mutants transplanted into hosts
double recessive for the same mutant and vermilion. The
results are parallel to those under 2.

4. Young Wild-type disks implanted in older vermilion hosts
developed intermediate pigment.

The first set of these experiments agrees with the assumption
that the Wild-type host furnishes the missing v+-substance
without which no vermilion develops. In the second group,
wherever the genetic color appeared, the implant must itself
contain the necessary v+-substance ; where, however, lighter
colors appeared, the mutant eye could not have a sufficient
amount of this substance. The third group shows the same
thing, but it is surprising that the double-recessive host does not
supply any v+-substance, which ought to be produced by the
Qonvermilion gene. It is probable that such experiments will
lead to further insight if put on a quantitative basis. The
present situation is conceived by Ephrussi and Beadle (19376)
in the following way: When a Wild-type eye disk is grown in a
mutant host and develops the mutant color, a substance is
supposed to be lacking or deficient that is needed for Wild-type
development. By checking different combinations against each
other, three such substances have been found which were called
v+ (missing in vermilion), cn+ (missing in cinnabar), ca+
(missing in claret). These substances are supposed to be formed

THE MUTATED GENE 167

as steps in a chain reaction in the order ca+ — + v+ — > cn+. ca+ is
assumed to be formed first because it is present in full concentra-
tion in flies lacking v+, cn+, or both.1 It reaches effective con-
centration before puparium formation; v+, however, only later.
V+ is supposed to be formed prior to cn+ because (among other
facts) a y-implant, deficient in cn+-substance if grown in a
cn-host (with the same deficiency) gives Wild eye, thus proving
its ability to produce cn+. In the production of these three
substances, a large number of genes is concerned. For the first
step no other gene but ca could be found to be effective. The
facts mentioned above for the transplants of many mutant disks
into v-hosts show that a number of these genes are concerned with
the production of at least the v+-substance, e.g., pn, pr, sf, etc.,
partly in the eye, partly in the body, e.g., v, p, rb. Cases in
which cn-implants are only partially affected by the mutant host
indicate genes concerned with cn+-production ; bri and mah are
mentioned as such. It is clear that thus far these relations
cannot be expressed in concrete terms. A general description
would probably again assume the form of the theory of reactions
in tune.

In other chapters, we reported the considerable amount of
quantitative work done with the Bar locus in Drosophila.
Hersh (1929), who pushed the analysis furthest, has also used
this material for an attack upon the problems of this chapter.
He took a selected homozygous stock of Bar, supposed to be
freed of heterozygous modifiers, as a basis and studied the effect
of other mutant genes in the X-chromosome upon facet number.
For obvious reasons (one X), the males only were used for the
counts. Hersh took into account also another set of facts, to be
reported below (page 215), viz., the different growth rate of the
dorsal and ventral eye lobes, which follow the equation for
heterogenic growth (see page 216) : Y = bXk, in which X and Y
are the facet numbers in the dorsal and ventral lobes; b the size
ratio of the lobes at the time of facet formation, i.e., the initial
difference; and k the ratio of the logarithmic rates of facet forma-
tion. The results of this inquiry are summarized (in simplified
form as compared with Hersh's table, 1929) in Table 20, which
compares facet numbers of a forked Bar line with those after

1 Recent work shows ca+ and v+ to be identical. The seat of production
of the substances is eye, fat body, Malpighiun tubules.

168

ruYsiouHihM. <;h.\/rncs

addition <>f the genes -cute echinus, crossveinless, cut, vermilion
garnet, and also the calculated values of b and A-

Table 20

From llcrsiii

Difference in

( renotype

facet number
from pure

b

h

fB line

sc

+

+

+

+

+

fB

+ 0.26

0.22

1.24

+

ec

+

+

+

+

fB

+ 4.87

1.05

0.85

+

+

cv

+

+

+

fB

+23.94

3.18

0.61

+

+

+

ci

+

+

fB

-15.28

1.24

0.80

+

+

+

+

v

+

fB

+ 18.51

0.24

1 .17

+

+

+

+

+

'.I

fB

-24.32

1.20

0.79

+

+

+

+

+

+

fB

0

0.37

1.11

sc

+

+

+

+

+

fB

+ 0.26

0.22

1 .24

sc

ec

+

+

+

+

fB

+ 5.54

0.50

1.02

sc

ec

cv

+

+

+

fB

+23.48

0.89

0.91

sc

ec

cv

d

+

+

fB

+52.18

2.28

0.71

sc

ec

cv

ct

V

+

fB

+58.06

2.14

0.73

sc

ec

cv

ct

V

g

fB

+ 2.73

1.95

0.75

+

+

+

+

+

9

fB

-24.32

1.20

0.79

+

+

+

+

V

9

fB

-17.11

0.49

1.07

+

+

+

ct

V

9

fB

-16.08

0.90

0.88

+

+

cv

ct

V

9

f B

- 7.65

2.49

0.66

sc

ec

cv

ct

V

9

fB

+ 2.73

1.95

0.75

This table shows that some mutant genes increase, others
decrease, the Bar action and are, in Bridges' terminology, plus
or minus modifiers. If more than one are involved, the action
might be cumulative or only of the same type as with one of the
members. The same gene may even act in one combination as
a plus and in another as a minus modifier. The values of b and k
are influenced in a similar way. There is no relation between the
normal, known actions of these genes and their actions in these
combinations. But the facts as a whole seem to me to indicate
that the reactions controlled by these mutant genes are primarily
of a quantitative type, shifting the velocities of some processes in
one or another direction, and that the time relations of these and

THE MIT AT ED GENE 169

the Bar reaction will influence the effect of the Bar reaction by
+ changing its time of onset, its speed, its end or threshold value.
In short, these facts again fit best into a general picture of the
type as expressed in the theory of reaction velocities in tune.

Interaction of genes, in the cases reported, was analyzed by
a comparison of the phenotypes, though the reactions leading to
these phenotypes were more or less inaccessible. In plants, an
analysis is available which is based upon the study of the chemism
of the process, viz., the work of Lawrence and Scott- Moncrieff
(1935) on flower color in Dahlia. We have reported (see page
95) the chemical side of this work. A special study was made
of the interaction of the different genes controlling the chemism of
flower color. (This is the work alluded to in connection with
dosage differences in polyploids, page 150.) The genes in
question are as follows : A is necessary for light anthocyanin color
produced by either cyanin or pelargonin. B is needed for heavy
anthocyanin pigmentation. I produces ivory fiavone and Y
produces yellow fiavone. As these Dahlias are tetraploid, each
factor is quadriplex. Y and B are completely dominant in
simplex condition; A is cumulative from simplex to quadriplex.
I is incompletely dominant ; when simplex, it produces very little
pigment; when duplex to quadriplex, the complete amount of
pigment. A fifth gene H acts as an inhibitor of yellow fiavone,
with a cumulative effect in the four quantities leading to cream
and primrose colors.

The interaction of these factors is, generally speaking, such
that the pigments suppress each other, but the flavones suppress
the anthocyanin more than vice versa. The degree of suppres-
sion depends upon the total number and the proportion of the
flower factors present. We may choose a few examples from the
large array of facts. Anthocyanin intensity (by A) is diminished,
when ivory fiavone (by /) is present, less pigment being actually
present. If the number of these genes is changed (1 to 4), the
effect is proportional; e.g., more /-genes increase the paleness
of the color by proportionally reducing the amount of antho-
cyanin formed. The relative effect of / is higher than that of
A, as the combinations of different quantities of both show. The
following grouping from low to high anthocyanin color (left to
right) in different combinations of A and / in 1 to 4 quantities
illustrates the point:

170 PHYSIOLOGICAL GENETICS

Less <>r More Ant hpcyanin
Little or No — »

.1,/ .»,/, A J, AJx AJ<

.1,/, -li/i A,7, .1:,/:, .1,/

.17, -I, AJ, AJ: AJ,

.1, -!:,/, A Ji

I .1,

Iii the same way, the other combinations were analyzed. }' acts
strongly upon both I and A, suppressing them completely in
some combinations; Y and I similarly suppress the action of B;
the action of / changes the effect of B to the type of an A effect.
Also, in these cases, the cumulative effect of the quantities of the
respective genes is observed. But it is not simple and appears
only in definite combinations, in which the potential cumulative
effect of a gene becomes an actual one; e.g., B is completely
dominant and has alone therefore no actual cumulative effect.
But when the B effect is changed by the presence of Y, the
cumulative effect appears. (This shows, of course, that a
threshold problem is involved, as in so many instances discussed
above.)

This potential and actual cumulative action as well as the
interaction in the form of suppression of the coproduction of other
pigments may be interpreted in terms of the chemistry of pigment
formation. (See page 96 for the basic chemical facts.) As a
balance of the different pigments is clearly involved, the authors
conclude that all pigments are being produced through some
common fundamental chemical reaction or from some limited
common source from which all pigments are derived and for which
they compete. (See the similar conclusions of Wright, page 157.)
The supply of this source must be so limited (see threshold,
page 158) that the quantity of each pigment can increase only
at the expense of the others, depending upon the relative quanti-
ties of the factors present. This hypothesis was tested by a
quantitative comparison of different combinations in extracts,
from which the relative demands of the factors upon the common
source was estimated. It was found, for example that in
BY -types both pigments are produced from a source that must be
three times as great as the maximum source in the A I -types but
smaller, when all four genes are present.

THE MUTATED GENE 171

On page 96, we mentioned the fact that anthocyanin may be
cyanin or pelargonidin. But in Dahlia, A and B do not control
one or the other; rather the whole quantitative system of the
genes involved determines this alternative. Pelargonin pro-
duction, for example, depends upon the following factorial
proportions :

1. A or B together with Y.

2. Two or more B.

3. One or more B together with three I.

4. One or more B together with two or more I and at least
one A.

5. One or more B together with one I and at least three A.
The whole body of these facts is summed up in a quantitative

representation based upon the arbitrary assumption that the
limit of pigment source that can be made available in Dahlia,
whatever factors are present, can be expressed as six units. The
potential units of the different genes are then calculated in the
following way : B and Y are capable of using the maximum source
in simplex condition and therefore contribute at least six potential
units. A has a cumulative potential value which in quadriplex
condition is less than or equal to the maximal source. Following
this way of estimation, unit values fitting the facts are fixed as
Y = 9, B = 6, 1=1, A = 1/2, and their combined action
would be, in a few examples, if 8 is the threshold value for
pelargonidin :

Only cyanin

AJt = 6 units)

BiAJi = 8 units \

A\Y\ = 9.5 units)

BiAJ2 = 8.5 units > Only pelargonin

Bo = 12 units )

Thus the entire body of evidence leads to the following con-
clusions :

1. The potential unit values of the four color factors are as
given above.

2. A sum total of six units is sufficient to produce the maximum
pigment source that plants carrying B or Y are capable of
supplying. In /U-plants, this maximum is two units.

3. Each dominant factor may contribute to this source accord-
ing to its potential unit value. But the actual units are limited
to 6 (or 2) whatever the potential units.

172 PHYSIOLOGICAL GENETICS

4. Bach factor competes for this source in terms of its potential
units. The total pigment production depends upon the propor-
tion and power of interaction of all factors.

5. V governs production of yellow flavone; 7, the ivory flavono
apigenin. // reduces the potential value of )'.

6. In the presence of A and B, cyanin is produced if the total
potential units are below 8; and pelargonin, if they are above 8.

7. Tentatively these actions in terms of units are related
to the constitution of the pigment molecule by assuming that they
are involved in controlling the oxidation of the phenyl ring.

There is no doubt that in this fine piece of work a considerable
step has been made toward interpreting the interaction of genes
in terms of the interplay of chemical reactions. In doing so, the
same types of principles were found at w-ork as in the other cases
that we reported : reaction velocities, proportional effect of gene
quantities, threshold conditions, competitive actions (see sex-
determining reactions, page 52). The system of genie action
here revealed is then the same as postulated in Goldschmidt's
theory of reaction velocities in tune, as the authors also indicate.

We may state finally that it would be desirable to have more
information regarding concrete types of reaction and the exact
nature of their interplay. What little is known in this respect
will be reported in a special chapter. Attention should be called
finally to some cases in which the disturbance of the proper tuning
of reaction can be described in relation to definite processes.
Goldschmidt has discussed, in a number of papers since 1920, the
many beautiful examples of this kind furnished by the facts of
intersexuality. The simplest and clearest is the following: The
end of morphogenesis of the organs of insects is given by the
process of chitinization of the soft organs, and normally this occurs
when development is completed. In intersexes, numerous organs
start development at a later date than normal, with the conse-
quence that at the time of chitinization a developmental stage
or even an imaginal disk is present where a completely developed
organ should be. In this case, the embryonic structure is being
chitinized and appears thus in the adult. Here the effect is
produced genetically by the improper balance of the reactions
controlled by female and male determiners together with the
unchanged general reactions of development. It is an actual
disturbance of tuning.

THE MUTATED GENE

173

There is another case that has also been frequently discussed
in Goldschmidt's writings — the abnormality called prothetely.
In these cases, which at least sometimes are caused genetically, a
single organ or a few organs are ahead in development. Most
typical are the cases in insects, where a caterpillar forms pupal
antennae (see Goldschmidt, 19236) or pupal wings (Tenebrio,
Ferwerda, 1928) (Fig. 33). As it is known now that the evagina-
tion of the imaginal disks is under hormonic control, this means
probably that in these cases the threshold has been lowered
locally. In any case, there is some dis-
turbance of reactions that are normally
definitely tuned and that have to do
with hormonic control of developmental
processes.

There can be no doubt, then, that the
combined action of the genes in control-
ling developmental processes occurs in
the form of a system of gene-controlled
reactions, which have to be in proper
tune to work together harmoniously. A
definite stuff, a definite physical or chem-
ical property of the cytoplasmic sub-
stratum, must be present at a definite
time and place to make the proper course
of other reactions possible. This inter- L
play of reactions in tune is the physio- , FlG- 33\7Case.,, of ,pTro~

*" f i • i • i thetely. Caterpillar of Ly-

logical process, which is meant (or OUght mantria dispar with pupal

to be) when one speaks of genie balance antennae. (From Goid-

. . schmidt.)

or of all the genes taking part in all

determinations. There is much loose thinking in this field.
One has, for example, argued as follows: There are so
and so many genes known to influence the eye color in
Drosophila. Therefore so many times more genes must be
involved in the determination of an eye, and so many more for
the whole normal development. The proper argumentation
would be: Numerous reactions are involved in orderly develop-
ment, including also the development of the eye and its details.
As the formation of eye pigment in definite cells is a chemical
process which is dependent upon a number of chemical and
morphogenetic situations, which in their turn are dependent upon

174 PHYSIOLOGICAL GENETICS

many others, it is to be assumed that mutations that change
reactions that have to do with any of these processes in the end
can also affect eye pigment. It might just as well be that 1 or
100 per cent of the gene-controlled reactions will have such an
effect, if tampered with by a mutation.

8. THE TYPE OF REACTION CONTROLLED BY THE MUTANT GENE

The facts thus far discussed point to the conclusion thai
mutant genes act by changing rates of processes concerned with
the harmonious progress of development. Many of the authors
quoted in this connection have tried to formulate concrete ideas
concerning these reactions. In so far as these ideas are of a more
speculative type regarding the nature of the gene, they will be
mentioned in a later chapter. In this chapter, we intend to
study those facts which actually demonstrate processes that
permit one to link the insight into the action of the mutant gene
with such insight as we have concerning the factors of development.

A. Reactions within the Cells

A priori, two types of reactions controlled by mutant genes are
imaginable, if wre do not consider the nature of the product of
reaction but only the dynamics of its production. One possibility
is that the gene-controlled process is confined to the cells in
which the mutant gene is situated. In terms of experimental
embryology, this would be a self-differentiative type of process.
The other type of reaction would be the production of substances
that spread from a center of origin over parts of the developing
embryo. In terms of experimental embryology, this would be a
process of the type of dependent or inductive differentiation. It
is very difficult to prove whether or not the first type exists at all.
The only possibility of attacking the problem in genetic experi-
mentation is the study of mosaics of genetically known constitu-
tion, which appear either as freaks or regularly in certain strains.
The work of Morgan and Bridges (1919), Patterson (1929a).
Sturtevant (1929), Demerec (1928), and others has furnished
important facts. If a Drosophila egg starts development as a
female, with two X-chromosomes, it was shown by Morgan and
Bridges (1919) that one of these X's is occasionally eliminated
from a cleavage nucleus. These cells will then be male, and the
result is a gynandromorph. (There are other modes of formation

THE MUTATED GENE 175

of gynandromorphs — Bee Goldschmidt, 193 Id — but they do not
furnish much information for the present problem.) If mutant
genes are present in the X-chromosome, the mosaic male spots
will also show the mutant character, and the facts showed that
the development of these is autonomous. Even very small
mosaic parts behave as if they were independent of the rest
of the animal. In these cases, then, the mutant genes must have
acted within the cells in which they were located, and their action
may have begun rather late in development. But it ought to be
emphasized that the characters involved here are such as are
actually formed only at the end of differentiation within the cells,
e.g., pigment formation or development of bristles. Develop-
mental processes that have to do with growth and differentiation
of a whole Anlage do not seem to react within the individual
cells; in mosaics, they appear only in a whole organ. This
applies, for example, to wing form.

These facts, derived from sex mosaics, reappear in cases where
mosaics have been produced in different ways. There are genes
like the Claret gene in Drosophila (Sturtevant 1929), or the
Minute-w gene (Bridges 1925), which produce elimination of
chromosomes and therefore mosaics. There is the possibility of
producing a similar result by inducing somatic mutations by X-ray
action upon individuals that have been marked by definite genes,
preferably by heterozygous sex-linked genes. A special study
has been made by Patterson (1929a) using eye colors in Dro-
sophila. The results coincide with those found in other cases of
mosaicism, viz., that mosaic areas may be formed from a whole
eye down to a single aberrant ommatidium. The genie effect
upon pigment formation then appears rather localized (see page
176 for more facts). Here it wras also possible to perform the
important experiment of producing the somatic mutation at
different times of development. The result was, as expected,
that the earlier the treatment the larger the mosaic area.

A third method of attack upon the same problem is furnished
by the study of what has been called unstable genes. In this
chapter, we are not concerned with the explanation of this
phenomenon but with the results. It is assumed that in these
cases mutant genes have the tendency to revert to the Wild-type
(or, more rarely, if ever, to another) allelomorph during develop-
ment. The result is a mosaic spot of different tissue, the size

176 PHYSIOLOGICAL GENETICS

depending upon the time in development in which the somatic
/Nutation takes place This phenomenon is rather frequently
found in plants, especially in regard to colors but also affecting
other morphological characters. Demerec (19356) counts 63
such characters in plants, especially studied by Emerson, Baur,
Demerec, and Imai (see Demerec 19356). In animals, only a
few cases are known, most of them reported by Demerec (see
Demerec 19356) in Drosophila virilis. Here a body color is
involved, and the miniature wing. (We mention the important
point — see page 212 — that the miniature wing is a complete wing
which did not finish the pupal phase by growth of the individual
cells, i.e., by a very late process in development.) In all these
cases, the existence of rather small mosaic spots shows that the
mutant gene acted within the cells in which it was situated.

There is a fourth type of evidence of the same order. Stern
(1936) studied mosaic spots that were produced by somatic
crossing over which occurred in small groups of cells. In this
case, also, very small spots were involved, which showed genie
action within individual cells of such an organ system as the skin.
Demerec (1934a) used a somewhat different method. He
also bred mosaic flies in which the mosaic spots were the results of
somatic segregation. These mosaic spots had been marked by an
ingenious method with small deficiencies (lack of a piece of the
chromosome at a definite locus); and as they arose by somatic
segregation, there were always adjacent twin spots with and
without the deficiency. Those with the deficiency will be
missing if the deficiency is lethal for the cells that carry it. The
size of the spot, of course, measures the time at which somatic
segregation occurred. Of 33 deficiencies in the X-chromosome
thus tested, all but one were lethal to the cells. This, of course,
shows only a general effect and does not contain informa-
tion regarding the time of action of a special gene-controlled
process.

These facts certainly show that many gene-controlled reactions
may be confined to a single cell, and it is possible that this type
of process is typical for some gene-controlled processes that take
place late in development and are localized in groups of identical
cells, without a pattern within this group. But even in these
cases a certain centrifugal action away from the cells containing
the mutant genes has been observed.

THE MUTATED GENE 177

Sturtevant (1927) produced mosaics of the Bar-eye mutant of
Drosophila with normal eye facets marked by color genes, by
introducing the Minute-n gene which has a mosaic-producing
effect, as mentioned above. He showed that mosaic spots of
Nonbar tissue situated in the periphery of the area of the would-be
normal eye are prevented from forming facets. Near the Bar
area, however, and in close contact with mosaic Nonbar tissue,
facets may be developed also in Bar tissue. The adjacent
normal tissue then exercises a facet-forming influence upon
tissue that would be unable to develop facets by the action of the
genes within its own cells. Sturtevant himself assumes that here
an influence beyond the cells containing the genes (an induction)
is seen, though he mentions also the alternative that a time rela-
tion might be involved, viz., a difference in regard to the time at
which the formation of facets is determined in different areas.
Huxley (1935c) has proposed a somewhat different explanation
in terms of growth rate, which however also requires a growth-
stimulating influence of normal on Bar tissue. (See also the work
of Huxley and Wolsky (1936) reported on page 33.) A similar
case is reported by Sturtevant (1932). He compared mosaic
patches of the mutant scute, which removes certain bristles in
Drosophila, with all-scute flies and finds that this scute effect
is smaller in little mosaic patches, indicating the possibility of
an influence of the normal surroundings on the mosaic spots.
Another similar case is found on a gray body with yellow mosaic
spots which are less yellow and less distinct than typical yellow.
Finally, a case ought to be mentioned that was described for
Habrobracon by Whiting, Greb, and Speicher (1934), and
A. R. Whiting, 1934. They produced eyes with mosaic spots
from binucleated eggs. Whenever any color from the orange
locus is associated in a mosaic with one of its alleles, no sharp line
of separation is found between the two, but continuous shading.
When, however, the white locus is involved, there is a clear-cut
mosaic. Adjacent cantaloupe and ivory tissue form a black
stripe in between, showing that from both spots reciprocal stuffs
diffuse which produce pigment when combined.

Of significance in this connection also are the remarkable facts
found by Haberlandt (1935) in the chimeric Crataegomespili,
which have a skin of one and a core of the other species. The
palisade tissue, derived from Mespilus, is intermediate in char-

178 PHYSIOLOGICAL GENETICS

.irtcr; and the Crataegus epidermis shows a clear influence of the
underlying Mespilus tissue. The structure is actually more
intermediate than in real hybrids. If from such a chimera pure
Mespilus shoots arc formed, containing only Mespilus cells and
chromosomes, an influence of the Crataegus genotype is visible
and persists. No doubt in this case morphogenetic stuffs,
controlled by the genes of one species, pass from cell to cell and
act upon the distant cells.

None of these cases is yet quite clear, and at present it may
suffice to register the facts.

There can be no doubt that some gene-controlled reactions take
place within the same cells in which the respective genes are
situated; furthermore, that this action may take effect even if it
is started rather late in development. It seems, moreover, that
this type of genie action is confined to such reactions as occur
either at the very beginning or at the end of development and
that in the latter case they lead to no further pattern formation.
In some cases, however, the products of these reactions may
diffuse into the surrounding cells.

In his wrork on sex, from which Goldschmidt originally had
derived his viewr on genie action, he thought that he was dealing
with a typical case of purely intracellular action of the genes in
question. Embryological and morphological facts which were
found later convinced him, however, in agreement with the
experimental work of Seidel (see Seidel, 1936), that in insects,
also, a part of the determinative reactions are not of the intra-
cellular, strictly self-determining type. We shall mention these
facts later. The foregoing discussion shows that such types of
action are to be expected wherever pattern formation is
involved.

It has to be kept in mind also that in experimental embryology,
the limits between self-differentiation and induction have turned
out to be rather vague', if not actually nonexisting, and that the
actual difference is a difference in time of irreversible determina-
tion. But there are also some cases existent in which the deter-
minative reaction, controlled by genes, actually takes place
within highly differentiated cells without induction from cells in
the neighborhood. Such cases may be analyzed in the mosaics
that the experimenter produces by transplantation as well as
in the genetic mosaics reported above.

THE MUTATED GENE 179

As an example of this type, Twitty's recent work (1936) on
newts may be mentioned. He transplanted the parts of the
embryo that produce the pigment cells reciprocally between
different species of Triturus, which are characterized by a different
arrangement of their pigment cells. The transplants kept the
arrangement of the species from which they came. An example
of the interactions of both types of differentiation is the well-
known experiments of Spemann, Schott6, Holtfreter, et al.
The differentiation of the region around the amphibian mouth is
of the inductive type. If the inductor is taken from a Triton,
and the tissue to be induced from a frog, the inductor induces
differentiations of mouth organs, but the specific type of organs
(horn teeth, etc.) is controlled by the genes in the induced tissue.

As far as experimental information goes, we may assume that
in vertebrates the inductor type is more frequent in development ;
in insects, however, self-differentiation is more frequently met
with, though not so exclusively as was believed a few years ago.
(Embryological work of Seidel and students; inferences from
genetic work as reported here.)

Thus far, we have considered only such gene-controlled reac-
tions within the cells as occur at the end of development. But
the majority of hereditary reactions occurring within one cell
are those occurring within the animal egg in its first stages of
development. In a later chapter, in which the problem of
pattern Avill be treated, we shall discuss the problems connected
with the early differentiation of the egg, and we shall mention
the cases in which genes are known to interfere in this process.
As yet there has been, in general, no genetic analysis of these
intracellular rearrangements of area with restricted determina-
tion (organ-forming stuffs, embryonic fields, etc.) which occurs
in the one-cell stage of the eggs of determinative type and con-
tinues through early development. The only point known is
that these processes are hereditary and specific and that they
may occur with different speeds in nearly related species. The
geneticist cannot doubt that such processes of distribution and
arrangement of cytoplasmic components occur under control of
the genes. The work of Haemmerling, which will be reported on
page 192, has proved such control at least for a complex plant cell,
and the author is in agreement with him when he extends his
conclusions to the developing egg. Experimental embryologists,

180 PHYSIOLOGICAL GENETICS

however, speak of the Bame processes usually as being of a cyto-
plasmic type. This is not correct, since all the facts indicate that
the same typo of processes may occur either later in development
or earlier and even in the undeveloped egg cell (see page 206).
This misleading description becomes dangerous when facts are
analyzed to show the part played by the maternal cytoplasm in
early development as compared with the influence of the genes con-
tributed by the sperm. If it is found that certain developmental
processes occur under complete control of the egg, this does not
mean that these processes are not under control of genes. It
means actually that certain genes present in the egg nucleus con-
trol the pattern formation in the egg and in the following stages.
In a hybrid egg, therefore, such a pattern ought to be of a hybrid
type, if the parents had different patterns. But this experiment
has thus far not been performed. As a type of analysis of the first
generation, we mention Hamburger's recent work on species
hybrids in amphibia (1936). He crossed Triturus crislatus, T.
taeniatus, and T. pahnatus and followed in detail the development
of the hybrids. In this case, the parental forms are different only
in the later embryonic stages, the tail-bud and early larval stages.
Up to a somewhat later stage, the limb bud stage, the embryos
are absolutely maternal in reciprocal crosses, and only later the
influence of the father becomes visible in influencing growth and
pigmentation. This shows, then, that a certain amount of
pattern formation before as well as after fertilization occurs under
the exclusive influence of the egg-nucleus genes. In terms of
intracellular gene action, it would mean that there is no difference
between the diffusion through the cell of substances produced
by action of the genes (with consequent arrangement as in a
pattern) and the diffusion of these substances from cell to cell.
Embryonic effects of genes and other effects of the usual type
are hardly different in principle (see also page 205).

We may point out finally that diffusion of gene-controlled
products from cell to cell may become visible in quite a number of
cases in which these products have only a general action. Baltzer
and de Roche (1936) discussed this problem in connection with
their experiments on growth of merogonic tissue upbn a different
host (see page 268). Here the otherwise impaired vitality of the
merogonic cells is reconstituted, and this might be interpreted
on a genie basis as a check to lethality by diffusion products

THE MIT ATED GENE 181

from the surroundings. But since Ephrussi (1935) has shown
that tissue containing homozygous lethals may be grown in
tissue culture, it might also be possible that an influx of some
product controlled by the normal chromosomes is not involved
in Baltzer's case but rather a draining off of poisonous products of
abnormal metabolism. In agreement with this interpretation
is also the fact that Hadorn (1934) was able to grow merogonic
cells in tissue culture.

B. Reactions of an Inductive Type

The experiments just mentioned lead to the type of embryologi-
cal process called induction. A number of facts are known that
prove that the products of gene-controlled reactions are formed
in definite areas of the embryo at a definite time and spread
hence into neighboring territories where they induce certain
morphogenetic processes. These facts constitute the most direct
link between genetics and embryology.

1. Terminology. — When Goldschmidt (19206) tried to derive
general ideas on gene action from his experiments in intersexuality,
he had to face the following situation. It was proved that sex
was controlled in Lymantria by male and female reactions of
definite velocity which are timed in such a way that one or the
other passes the threshold. Since all the facts of gynandro-
morphism seemed to prove that insect development is of the
completely self-determinative type, it had to be assumed that the
sex-controlling reactions take place within each cell, producing
sex-determining stuffs. It was also known that in vertebrates,
morphogenesis of sex differentiation is controlled largely by sex
hormones; hence it was concluded that these intracellular sex-
determining stuffs are physiologically identical with the sex
hormones. The difference would be only the production within
all body cells versus production in a centralized organ, the gonad.
From this, the generalization was derived that all genes produce
substances of morphogenetic action similar to hormones, and,
therefore, the general term hormones was amplified in its scope
to cover also such intracellular morphogenetic stuffs. When
later Witschi (1929) proved that in vertebrates a special sex-
determining action was exercised by cortex and medulla of the
gonads whence it spread by diffusion, Goldschmidt (1931c?)
proposed to distinguish three types of morphogenetic substances

182 PHYSIOLOGICAL GENETICS

produced by the action of genes: hormones of the first order —
intracellular formative or inductive substances; hormones of
the second order inductors diffusing from their place of origin
and thus acting at a distance (identical with the harmozones of
Gley); hormones of the third order, or hormones proper —
carried by the body fluids. Meanwhile, the work of the embry-
ologists has centered considerably around the second type of
inductors; and it is necessary to use a definite terminology for
the possible morphogenetic stuffs. That the term hormone was
extended beyond its original meaning to cover all the types of
stuffs has been much criticized. I still think that the general
idea underlying this generalization is correct; but this does not
prevent the use of any terminology that appears better founded.

In a recent paper, Huxley (1935a) proposed a clarification of
the terminology. He points out that all transitions are found
between the types of morphogenetic substances mentioned;
e.g., the activating substance studied by Haemmerling in the
unicellular plant Acetabularia is, of course, intracellular, but it
is shown to act by diffusion from the nucleus to the mushroom-
like periphery of the complicated cell. Similar examples are cited
that bridge the difference between diffusion hormones and vas-
cular hormones. He proposes, therefore, to call all these sub-
stances activators. They may be local activators acting on the
cell or tissue that produces them. These again are subdivided
into intracellular activators and chemodifferentiators, the latter
containing all the activators for predetermination of embryonic
parts. (These may also show transitions to a type with diffu-
sion.) The second main group is distance activators, or hormones.
They are transported either by diffusion — diffusion hormones —
or by the body fluids — circulatory hormones. This system is,
after all, very much like that used by Goldschmidt. We shall
then use the word activator in a general sense, replacing it occa-
sionally by inductor or evocator when the embryological processes
of a general type are involved ; and we shall use the word hormone
mostly for distance action only.

2. Genie Products of the Hormone Type. — A few facts are
known that might be reported as a transitional step between
the cases with a spread of a gene-controlled effect from cell to
cell and a corresponding effect at a longer distance. Bonnier
(1928) described a Drosophila mosaic in which a Bar eye on one

THE MUTATED GENE 183

side made a Wild-type eye on the other side smaller (if this is
the correct interpretation).

But the numbers of cases are increasing in which clear relations
are found between genie action and products that act as hormones
Probably the first case outside the sphere of sex hormones was the
one discovered by Sturtevant (1920) in the case of vermilion,
a gene for eye color in Drosophila simulans. Sturtevant produced
gynandromorphs in which the gonads were always vermilion
in constitution if they were testes and not vermilion or Wild
type if they were ovaries. When such individuals had mosaic
spots of vermilion constitution in their eyes, these spots were not
vermilion if ovaries were present. The Wild-type ovary then
suppresses the action of the vermilion gene in the eye, or, expressed
differently, the Wild-type gene present in the ovary makes this
organ produce a hormone that acts upon pigment formation in
the eye. (Another eye color, garnet, was not influenced by such
a hormone.) This case is closely related to another one found by
Dobzhansky (1931): white eye color in Drosophila is associated
with a colorless sheath of the testis which, however, is yellow in
Wild-type animals.

In gynandromorphs having the male tissues white, testes and
vasa efferentia are transparent in young individuals, but these
parts become yellow with age. If, however, an ovary is attached
to male organs, the adjoining genetically white parts of the vas
efferens soon become yellow. Dobzhansky thinks that the Wild-
type gene leads to the production in all tissues of a pigment-
precursor substance which might eventually reach the vas
efferens which is able to convert it into pigment. From the
Wild-type ovary, however, this substance may diffuse directly
into the adjacent vas efferens.

A similar case was also found by Whiting in Habrobracon
(1934). Genetically ivory mosaic spots in eyes show orange
color if the gonad of the mosaic is genotypically black.

A comparable case in the flour moth, Ephestia kuhniella, has
been worked out in considerable detail by Caspari (1933, 1935,
1936), Plagge (1935, 1936a, b, c), and Kuehn et al. (1935, 1936).
It furnishes thus far the best known example of a gene-controlled
hormonic action in development. There is a recessive eye-
color mutation aa making the otherwise black eyes red and,
simultaneously, the dark testes white. Caspari (1933) made

ISA PHYSIOLOGICAL GENETICS

transplantations of testes with different combinations of the
two races. When an AA-testis is transplanted into an aa-
host, it remains pigmented, and the eyes of t lie host turn black;
the result is thru the same as in the two Drosophila cases com-
bined. (A transplantation is, of course, another method of
producing mosaics.) This primary experiment was then followed
up by Kuehn, Caspari, and Plagge (1935). The genes aa affect,
in fact, quite a number of characters in addition to eye color and
gonad color: the skin of the caterpillar is paler; the color of the
optic ganglia is more reddish (as opposed to brown); the amount
of pigment in the larval ocelli is changed; the velocity of develop-
ment and the vitality are decreased. This shows that the gene
in question acts all through development. For one character,
the pigmentation of the ocelli, it can be shown that a different
velocity of pigment formation is involved. Probably the same
applies also to the color of the ommatidia, which shows in
oo-individuals less and lighter pigment and smaller granules.
(There is also an intermediate allele ak with intermediate effects
existing.) As Caspari found, aa-testes transplanted into
A A -animals became dark; AA-testes in aa-animals remain dark.
AA-testes in aa-hosts make their eyes darken; but the result is
not exactly the same as in pure A A -eyes; the quantity of pig-
ment remains less than in the AA-type, and also the distribution
is somewhat different. Also, if these transplantations are made
sufficiently early, the larval pigment characters are influenced
correspondingly. The A A -testis then produces a substance
(or hormone) which acts all over the body. The testis is, how-
ever, not the only organ that produces this substance : implanted
ovaries act like testes, though less strongly; by implantation of
more than one ovary the effect is increased. The same effect,
but still weaker, may even be produced by transplantation of
brain. Most remarkable is finally the fact that this hormonic
effect may act upon the egg cell even before fertilization: if
aa-females contain transplanted AA-testes, their pure aa-offspring
may show the early pigmentation of the ocelli characteristic for
A A. The hormone then has entered the cytoplasm of the egg
within the ovary. (We may point out that here the procedure
has been reproduced that must be assumed to take place in the
cases of so-called maternal inheritance of Toyama (see page 206).
Especially does this last experiment show that the stuff produced

THE MUTATED GENE 185

in the testis or ovary or brain by the action of the gene .1 is
actually a hormone. Whether this hormone acts also within the
cells in which it is found has not yet been decided.

To these facts have been added recently a number of quantita-
tive data by Kuehn and Plagge (1936). The presence of an
A -testis in an a-pupa for only 24 hr. suffices to produce the effect.
One grafted ovary produces less of the hormone than one testis
in 24 hr. ; two ovaries produce more. A single egg tube, which
is dissolved after transplantation, produces more hormone than
a whole intact ovary, viz., about the same amount as a day's
production of a testis. (This is proved to be actually boun<$ to
the presence of the .4-gene.) A dissolved testis, however, does
not produce the hormone. It seems therefore that the testis
produces, and the ovary stores, the hormone. The same hor-
mone occurs in many other species of Lepidoptera, as trans-
planted testes from many different species produce the A -effect.

Another method, invented by Beadle and Ephrussi (1935-
1937), has led to results that fall in line with those already
mentioned. (Part of their work has already been mentioned
on page 166.) They transplant in Drosophila the imaginal disks
of eyes of different genetic constitutions into the body of larvae
of definite constitution and recover the transplant after it has
finished development in the host. Three types of result are
possible. Either the transplant is not influenced by the host;
i.e., its development is autonomous; or the transplant assumes
the color of the host. In the latter case, the host must produce
a hormone-like substance, acting upon the transplant. As a
third alternative, the transplant may be intermediate in color,
which would amount to an insufficient hormonic action of the
host. The diagram (Fig. 34) shows the authors' results with a
large number of eye-color mutants. The majority of eye colors
show autonomous development. But vermilion and cinnabar
are exceptions; they are influenced by the host. Vermilion
implanted into Wild type or into all the color mutants marked
on top of the diagram gives Wild-type eyes with a few exceptions:
it gives vermilion in claret (ca), carmine (en), peach (pp), and
ruby (rb) and an intermediate condition in carnation (car) and
garnet (g2) hosts. Parallel are the results with cinnabar, as
may be read from the diagram. A Wild-type disk transplanted
into mutants always gives Wild-type pigmentation with the

ISC,

PHYSIOLOGICAL GENETICS

exception of a transplant into a ca-host , which becomes claret.
There is finally another case, in which the transplant changes
the host: a cinnabar eye disk transplanted into a host that is
simultaneously apricot and vermilion makes the host's eye

+

Host

+ bobwca\carcd cl rm\cn n- Hn 1 1 mapr pd'pn pr ras

6®o mo do oooooo'oo o'o'o

rb \se

O'O

fed

o

sf*\st V w wa w

oood

bo

©o

bw

o

o

ca

o

o

o

car

o

o

—

cd

o

o

o

o

o

o

cl

o

o

cm

o

o

cn

•

•

•

o

A

•

•

o

o

A

•

•

•

o

•

•

•

•

o

•

•

•

•

o

•

92

o

o

+- Hlf

o

©

c It

o

o

° ma

o

o

c.PP

o

o

£ pd

o

o

pn

o

o

pr

o

o

ras

o

o

rb

o

'd

se

o

o

,

sed

o

o

sf2

o

o

St

o

o

o

o

o

o

■annn^^nnHnHnnnsnnnn^nnnnannn

w

o

wa

o

o

o

we

o

Fig. 34. — Diagrammatic representation of the first series of results of eye
transplantation in Drosophila. Shaded circles, autonomous development of
pigmentation of the implant; black, nonautonomous; half shaded, intermediate
color. {From Beadle and Ephrussi, 1936, Genet. 21.)

apricot. Thus, we have "hormonic" influences from host to
implanted eye and from implanted eye to host's eye in certain
cases. The first influence, however, did not originate in the gonad
(as in the cases reported before) but in the fat body and Mal-
pighian tubules.

From these results the authors conclude: Since the pigmenta-
tion of a genetically p-eye can be modified to v+ (Wild type) by

THE MUTATED GENE 187

transplanting it to a Wild-typo host, the host is supposed to
furnish a substance (or hormone for its production) absent in a
j-fly, a #+-substanee. The same argumentation applies to a
cn+-substance, which is proved to be different from the ^-sub-
stance, because a y-disk in a cn-host gives a Wild-type eye. In
the Wild type, these substances are produced within the eye.

Recently Ephrussi, Clancy, and Beadle (1936) demonstrated
the presence of this stuff in the lymph by injecting Wild-type
lymph into apricot-vermilion pupae. The flies hatched with
apricot eyes. It turned out that the stuff is produced in pupae
30 to 80 hr. old and acts upon pupae from before pupation up
to 70 hr. after pupation.

The results regarding the claret character had shown that a
genetically Wild-type eye could not become Wild type in a
claret host; there must also be a ca+ substance, necessary for
Wild type, which, however, is not formed in the eye itself.
But as a ca-type eye in a Wild host does not become Wild type,
there must also be something in the ca-effect within the eye that
prevents Wild-type pigmentation. These three substances must,
however, be closely related because the absence of one usually
entails the absence of all. From these and other facts the authors
conclude that the three substances are successive products of a
chain reaction: ca+ — > v+ — * cn+. The different mutant genes
then produce either an interruption of this chain at some
point or a retardation of its rate. In this way — which links
these facts with our former discussions on rate — the details
of all experiments may be explained. (See, however, foot-
note, p. 167.)

Recently (1937), Medvedev has used the Ephrussi-Beadle
method for transplanting wing disks of mature larvae belonging
to Wild type, yellow, ebony, and black into different hosts in
different combinations. The result was always autonomous
development, which means a negative result from the standpoint
of this chapter.

At this point also an example from the plant kingdom may be
inserted. Anderson and de Winton (1935) in studying the
manifold action of some genes in Primula found that the genes
Ch increase in the flower sepal and petal lobing; the gene t
inhibits it; and both have an inhibiting effect upon peduncle
elongation, though at opposite ends of the peduncle. They think

188 PHYSIOLOGICAL GENETICS

that this points to a difference in distribution of some regulatory
substance, possibly a growth hormone. There is also a third
gene n which probably nets upon the production of the same stuff.

In this chapter, we ought to mention also those cases where
actual and known hormones are controlled by genie action.
[We have discussed (page 77) the facts of inheritance of the
number of molts in insects, controlled by a series of allelomorphic
genes. In these cases, however, we need not assume that the
genes in question lead to a production of hormones but only that
they are connected with the time of release of a hormone.
For this reason, the case was reported in the chapter on rates.]

There is, for example, the case of the dwarf gene in mice as
analyzed by Smith and McDowell (1930). It was shown that
this gene works by controlling the production of a growth hor-
mone in the pituitary gland, the insufficiency of which leads to
dwarfism and other abnormalities. Another probably more
complicated case taken from the sphere of sex hormones is the
much discussed case of the hen-feathered Bantam cocks. Hen-
feathering is controlled by a dominant gene (perhaps two), as
found by Morgan (1919) and Punnett and Bailey (1921).
Castrated cocks become normally cock-feathered, and trans-
plantation of any testis produces a return to hen-feathering. In
normal fowl, the male type of plumage is the genetic type, and
the female type is caused by the female sex hormone. According
to Callow and Parkes (quoted from Haldane, 1935), the Bantam
feathers respond with female type to a very low concentration of
female hormone. As there is some female hormone also in normal
testis secretion, the gene responsible for hen-feathering produces
a low threshold of response to the female hormone. If these
results are as significant as they appear, the gene in question
actually does not control the production of the hormone, but a
condition in the feather follicles, setting a threshold to hormone
action.

It will probably be found that the production of hormone-like
substances, i.e., diffusion and vascular hormones, plays a much
more important part in development than is known at present.
We expect that such processes will be found wherever an organ
differentiates in some respects as a unit. We know already some
cases from which such a situation may be inferred. There
is the case of the insect gonad, as analyzed in the intersexuality

THE MIT AT ED GENE 189

experiments, i.e., in a case where the normal genie action has
been replaced by another one. Goldschmidt (1931c) studied
the details of the transformation of a testis into an ovary, and vice
versa. He found that the facts were not in favor of the assump-
tion that the determining processes went on in the individual
cells. He made it probable that the balance of the sex genes in
this case either stimulated the sheath of the ovary to produce a
hormone acting upon the whole gonad in transforming it into a
testis; or, on the contrary, stimulated a group of cells in the
attachment point of the vas deferens to the testicular compart-
ments to produce a hormone, transforming the testis into an
ovary.

From the study of intersexuality in Lymantria, many more
cases may be derived. It is frequently found that an organ that
has been developed first male and ought to become female later,
or vice versa, after the turning point continues with its male
differentiation, when that has once started, if no female homolo-
gous part exists into which it could be transformed. For
example, in male intersexuality the valvae and penis of the genital
armature may fully differentiate after the turning point; whereas
the uncus is transformed into its female homologon, the labiae.
From this we may conclude that at a certain stage of development
the whole Anlage is determined by a hormone-like substance,
produced by genie action; furthermore, that the action of this
hormone has the maximum effect after a definite time of activity,
just as in Kuehn's experiments (page 185) 24 hr. presence of
hormone producer sufficed for full effect. It might be mentioned
incidentally that here also is found the explanation for the special
type of intersexes with both ovaries and testes produced in
Drosophila by Lebedeff (1934). A similar conclusion may be
derived from recent experimental work by Bytinsky-Salz (1933)
and Bodenstein (1936). Salz found that certain species hybrids
of Lepidoptera are unable to develop beyond a certain stage.
The same organ transplanted to a pure species will be induced to
behave like the host's organs. This looks like a kind of hormonie
control of differentiation.

From the standpoint of experimental embryology, the facts
just mentioned would be called an induction, and the application
of the hormone concept is not absolutely necessary though rather
probable, if cases like the gonad transformation are studied.

190 PHYSIOLOGICAL GENETICS

The specific struct me of the testis, for example, makes an induc-
tion by simple contact rather improbable. After all, the differ-
ences between induction by diffusion and by hormones are found
to disappear the more we learn about induction. Two sets of
facts demonstrate this. Greenwood and Blyth (1935) showed
that the sex hormone in quantities below a certain threshold is
not distributed by the blood stream but diffuses only from cell
to cell; the same substance is then once inductor by diffusion
and again a real hormone. Another example relates to the other
end of the series: Goldschmidt (1921a) found that in intersexual
males of Lymantria, frequently duplications and triplication
of parts of the male genital armature (the valvae) appeared.
He concluded that the cause must be the increased breadth of
the region of formation of these organs caused by the larger size
of the intersexual abdomen. Recent work by Holtfreter (1936)
shows that this explanation was right and why. This author
showed that a stretching of the amphibian organizer resulted
in multiple inductions. We concluded from the facts mentioned
above that an induction of the hormonal type must be present
in these parts of the genital armature. Here we have another
proof, which in addition links the facts with the typical cases of
embryonic induction.

3. Diffusion of Growth- or Similar Substances. — One of the
most frequent actions of the mutant gene is the production of
insufficient quantities (as compared with the normal gene) of
substances needed for normal differentiation. Lacking detailed
information, we might generally call them growth substances.
We have met them before in cases where possibly processes within
the individual cells were involved, e.g., the bristle-forming mate-
rials in Plunkett's study (1926). A case where we can hardly
escape an interpretation in terms of production of such sub-
stances, or their reciprocals, lytic substances, destroying previous
growth, was the scalloping effect of the vestigial series, and the
interpretation was already given in this sense. Here there can
be no doubt that the mutant development is either short in some
substance necessary for normal growth and differentiation;
furthermore, that this deficiency may be of graded type and
becoming effective at different times of development in different
alleles or compounds; furthermore, that the insufficient quantity
in question spreads across the wing from its base and reaches

THE MUTATED GENE 191

the tip last; or that the diffusing substance is a lytic substance
which is first accumulated in a quantity above the threshold of
action in the wing tip with all the other graded consequences.
The action of many other genes is of a similar type; the truncate
wing and the miniature wing in Drosophila are examples in
which other substances are involved, in the former case acting
only upon mesoderm differentiation; in the miniature case, only
upon growth after differentiation. This type of growth sub-
stances, in insufficiency and in hyperproduction as well as
eventual lytic substance, might be responsible for a large majority
of developmental processes in mutants, especially wherever
linear growth or reduction is involved. They might also be
responsible for all processes of differential growth; but here, of
course, the pattern factor will come into play, perhaps in the form
of different resistance to the diffusion of the substance in differ-
ent directions of space. This part of the problem will be treated
in a later chapter. It must be emphasized that in all these cases
the unity upon which the mutant gene acts is a complicated
system. The unit affected might be the whole organism (size
and proportions) or a whole organ (an imaginal disk, a wing)
or a definite part of an organ (the wing veins as a whole or one
definite vein). Within these units, therefore, no mosaic forma-
tions are possible except when the last stages of development are
involved (see miniature wing, Demerec, page 176). Moreover,
different processes of this type, conditioned by different mutant
genes, controlling the production of different substances, might
be combined in the same individual without interfering much,
with each other. For example, Goldschmidt (1937) showed that
the large broad wing of the mutant expanded in Drosophila is
determined even in the stage of the imaginal disk. A combina-
tion of the mutant genes expanded, dumpy, cut can produce a
wing that is short-dumpy, broad and cut, in addition, because
three substances are involved, produced at different times and
acting upon different tissues, viz., upon the general growth and
upon epithelial and mesodermal differentiation (for data see
Csik, 1933; Goldschmidt, 1937). If the same reactions with the
same end product were involved, a compromise result, e.g.,
intermediate, would be encountered. This, of course, is the case
with compounds of multiple allelomorphs; but it might also be
the case with different <rciics. e.g., different genes for scalloping,

192 PHYSIOLOGICAL GENETICS

like beaded and beadex.1 This problem has already been
discussed in the chapter on interaction of the genes (page 161).
Of a different type bu1 very important in connection with the

problems of this and the last chapter is (he work of Baemmerling
(1934) on Acetabularia (See also page 179). This alga has the
form of an umbrella with a long stem growing from a root like
base. The umbrella-like spread is formed after a series of wreaths
of hairs have been formed at the distal end of the stem and have
disappeared. The umbrella itself consists of a number of indi-
vidual radial chambers, in which the gametes will be formed later.
This whole plant is only a single cell with a large nucleus situated
at the base of the stem in the rhizoid. By cutting the stem, a
uucleated part containing the rhizoid and an anucleated one
ending in the umbrella, or "hat," will be obtained. The
anucleated fragment will live a very long time but is finally
doomed. Such a nucleated segment easily regenerates the lost
parts. But an anucleated fragment may also regenerate the hat
at the anterior as well as at the posterior cut end. It was then
shown that this regeneration of anucleated fragments takes place
more readily at the anterior end and that its success depends upon
the presence of a substance that is used up during regeneration
and shows a gradient of concentration, increasing toward the
end of the stem. The same piece will also regenerate a rhizoid,
and the stuffs responsible for this show the opposite gradient
with a maximum concentration toward the base of the stem.
The influence of the nucleus could then be tested in the following
type of experiment. A piece with a nucleus was allowed to
begin regeneration, and after it had been started the nucleus
was removed. The piece now left had complete power of regener-
ation though derived from the part of the stem that otherwise
would not form a hat. The nucleus then had produced the
formative stuffs which spread along the stem. Such experiments
show the nuclear origin of the formative stuffs but do not prove
that it is the genie material within the nucleus that is involved
here. This problem was attacked by using a second species
(1. A. mediterranea, 2. A. Wettsteinii) , differing from the first
mainly by the number and shape of the chambers constituting
the hat, the color, and a number of other characters. It was
possible to transplant anucleated fragments of one species to
1 Unpublished work.

THE MUTATED GENE 193

nucleated ones of the other in both directions. In this case, the
nucleated part controlled the differentiation: a mediterranea
stem upon a nucleated Wettsteinii rhizoid forms a typical
Wettsteinii umbrella. This shows that actually the genes
within the nucleus control the production of a specific formative
stuff (not unspecific, as in the hormonic type) which diffuses
through the cytoplasm to the place of its form-controlling
action.

It might be added finally that Gregory and his collaborators
(1933-1936) have tried to trace the actual product of genie action
responsible for hereditary size. As reported before (see page
54), hereditary size differences in mammals and birds (Castle
and Gregory; Gregory et at.) as well as in insects (Goldschmidt)
are based upon a different rate of cell division from the very
start of development. According to a number of physiological
investigations (Hammett, 1931), the concentration of gluthatione
in the body is somehow correlated with cell proliferation. Greg-
ory therefore tested the gluthatione content of embryos of large
and small breeds of rabbits and of fowl and found them propor-
tional to the adult size of the race. As it is believed that a high
concentration of disulfides favors protein synthesis, the size genes
might act via such intermediate steps.

4. The Determination Stream. — There is one series of facts
that allows the forging of another interesting link between genetics
and experimental embryology. When Goldschmidt (19206) first
developed his theory of gene action, he considered certain cases
of pigment patterns in Lepidoptera with which he was familiar.
There were especially cases of melanism which he had analyzed
genetically, and which showed that the amount of pigment
production, and also the arrangement of this pigment, was
dependent upon a number of Mendelian genes. In some cases,
it seemed that pigmentation started from one point (the wing
base) and spread from there in different genie combinations
across the wing in a definite path. In others, there seemed to be
a number of predetermined centers at definite points from which
this flow took place. Therefore the terms outlet of pigment and
-pigment stream, were used, a movement not of pigment, but of
pigment-determining substance being meant.

Later, another remarkable case was studied which led to a
more definite statement (Goldschmidt, 19236). Intersexual

194

/'// YSIOLOOH 'AL 0ENETIC8

males of Lymantria dispar are genetic males which have developed
up to a certain time as males but, from a certain point on, have
become female. Higher degree of intersexuality mean.- earlier

Fig. 35. — Scries of intersexual males of Lymantria dispar.

turning point.1 In this case, the wings exhibit a characteristic
behavior (Fig. 35). In the very first stages of intersexuality,
only one wing may show a small patch of female color and scale
structure. (The problem of symmetry involved here will be
1 In a scries of recent papers Baltzer and Kosminsky have attacked the
concept of a turning point at different times. A paper in press by the
author disposes of their arguments in detail.

77/ A' M ( 'T.\ TED GENE 1 95

studied in a later chapter.) With increasing intersexuality,
these patches increase on both wings without being symmetrical
and finally cover a whole wing. If we look at such a series of
wings of intersexual males of increasing femaleness, we are struck
by the fact that the arrangement of the dark male and white
female parts, irregular as it is, is of such a type that one might
describe it in terms of a stream of pigment flowing from the wing
base over the wing along the veins, covering the surface in an
irregular way, leaving islands in between until the stream is
stopped when all available pigment has flowed out (more in lower,
less in higher intersexuality). Obviously, the quantity of
"pigment" is a definite one for each stage of intersexuality and
available simultaneously to all four wings, to which it is dis-
tributed by chance. An exact study of numerous individuals
(Minami, 1925) bears out the general correctness of this picture.
Now, within dark male areas the form and arrangement of the
scales is of the male type; and within the light areas, of the female
type. The substance that has flowed across the wing therefore
determines color, shape, and arrangement of scales of male type.
As the amount of this substance is proportional to the degree of
intersexuality, and as the degree of intersexuality is proportional
to the time of occurrence of the turning point, it follows that the
simplest picture that we can make of the process is the following.
The differentiation of the male-type wing scales, in all their
features, is controlled by a determining stuff, which enters the
wing Anlage at its base and slowly flows across the wing, the
channels being determined by the actual structure of the wing
at this time. If the male is intersexual, this flow ceases at a
definite time which is controlled by the degree of intersexuality;
and the result is that all parts of the wing that have not yet been
reached by the flow will develop as female parts. (The details
may be conceived in different ways; the problem, however,
belongs to the problems of sex differentiation). From this it
follows that the visible patterns of male color and structure in
a series of male intersexes indicate the progress and type of flow
of an underlying determining substance. Adopting a conception
introduced by Spemann, we call it the determination stream
which thus is made actually visible; and inasmuch as the link
with the specific genie combinations that produce intersexuality
is known, we see that genes actually controlling color and struc-

196 PHYSIOLOGICAL GENETICS

t urc of a wing may act by controlling a determination stream of

definite quantity, speed of progress, pattern of flow, and action
upon different processes of morphogenesis and chemism. It is
surprising that these points of primary importance for the prob-
lem of gene action have never impressed other workers (with the
exception of J. S. Huxley).

There is another point of greatest interest in connection with
the same work. II a male Lymantria dispar becomes intersexual,
the female wing color and wing structure appears upon the male
wing, as just described. If, vice versa, the female becomes
intersexual, the male wing color appears upon the female wing
at once, all over the wing. It is known that the pattern of wing
color is determined in a sensitive period before pupation, and we
have seen how in the male the determination stream flows across
the wing. The different behavior of the intersexual female must
then mean that the time relations in regard to the flow of the
stream across the wing are different in the female; even the
latest turning point could leave still enough time for the determin-
ing process for the male color (and structure) to spread over the
entire wing. (Actually, all time elements of growTh and differ-
entiation are different in both sexes, as analyzed in detail in
Goldschmidt 1933.)

Giersberg (1929) has described and pictured a number of
intersexes in different Lepidoptera which in part demonstrate
beautifully the conception of the determination stream across
the wing, an interpretation that he also accepts.

In recent years, Henke and Kuehn have applied the same con-
ception to the interpretation of their experiments in shifting the
wing pattern in Lepidoptera. The pattern aspect of the problem
will be discussed in a later chapter. Here we consider only their
proofs for gene action by controlling a stream of a determining
stuff. We refrain from reporting Henke's first experiments on
Saturnid moths, because no genetic work is available there, and
report only the work on the flour moth Ephestia kuehniella
where genetic, phenocopic, and experimental evidence are
available.

Since the work of Sueffert (1925, 1927, 19296), Schwanwitsch
(1928-1929), and Kuehn (1926), it has been known that the com-
plicated pattern upon the wing of Lepidoptera is the result of
different arrangements of a number of always present constituent

THE MUTATED GENE

197

parts or fields. One such characteristic field (see Fig. 36) is
the symmetry field in the center of the wing, limited on each
side by a system of bands. Its unity is proved by the fact that
it acts as a whole when changed in phenocopic experiments and
that it is affected as a whole by mutant genes. Definite tempera-
ture treatments may broaden this field if applied at a sensitive
period or else narrow it. The same action is produced by a
mutant gene Sy, which constricts this field. To find out how the
size of this field is controlled, Kuehn and Henke (1936) (see also
Kuehn and Engelhardt, 1933) destroyed definite parts of the

Fig. 36. — Diagrammatic representation of the elements of the pattern in
nymphalids. Interpretation: left, by Schwanwitsch; right, by Sueffert. (From
Henke, 1935, Verh. deutch. Zool. Ges., Fig. 15.)

developing wing at different times and watched the effect with
the following result: If the wound is made at the proper time
(early pupa), the bands limiting the symmetry field are pushed
around the wound in a definite way, which is well illustrated in
Fig. 37. The resulting pattern is such as would be caused if a
determination stream proceeded from the anterior and posterior
margin of the wing in a broad front toward the center and if the
edge of this stream, whever it comes to a standstill, determined the
position of the bands limiting the symmetry field. A comparison
of Fig. 37 with the diagram (Fig. 38) beautifully illustrates the
actuality of this conception. It is clear that earlier operations
will produce this edge effect nearer to the wing margin and later
ones toward the center and that still later ones might only affect
the contour of the bands which become transverse after the two
streams have united (see Fig. 37). Here, then, the embryological

19S

PHYSIOLOGICAL GENETICS

experimentation demonstrates exactly the same thing as was
derived by Goldschmidl from genetic experimentation.

Furthermore, in this case, a number of facts may bo coordi-
nated.

1. The symmetry field may
be enlarged and narrowed by
temperature shocks at definite
and different sensitive periods
(sec Fig. 7, page 21).

2. The same effect may be
produced by a mutant gene.

3. The shifting effect upon
the bands by destruction of
wing tissue can be obtained
best at the time of the sensitive
period for narrowing the sym-
metry field (48- to 60-hr.
pupa). Heat, as well as the
gene Sy, acts by affecting quan-
titatively the course of the
determination stream. This
course may be directed by con-
trolling the quantity of the
determining substance, the time
of its flow, or the limits that
may prevent further expansion
(which Goldschmidt had called
the conditions of the system).
Kuehn and Henke speak of force
of expansion and force of resist-
ance (of and to the stream)
which is the same as quantity
or time of flow of the determin-
ing stuff and the conditions of
the system in Goldschmidt 's old

terminology. It might finally be added that Kuehn and
Engelhardt (1936) performed the same type of experiments upon
the currant moth, Abraxas grossulariata, with practically the
same results aside from minor differences in detail.

Fig. 37. — Right pupal wings of flour
moths, black race, operated in pupae
(0 to 24 hr. age). (From Kuehn and
Henke, 1936, Abh. Ges. Wiss. Goettingen,
15, Fig. 96.)

THE MUTATED GENE

199

The idea of gene action by means of a determination stream has
also occasionally been employed by Drosophila workers. Thus
Plunkett (1926) found that the bristle-destroying action of the

Fig. 38. — Flour moth, a, diagram of the spread of the determination stream;
h I, individual right pupal wings; b-f, black race; g-l, wild race;/, /, controls.
Age at operation g below 6\ b below 24fc c-e, h-k 24-72A. (From Kuehn and
Henke, 1936, Abh. Ges. Wiss. Goettingen 15, Fig. 97.)

achaete gene spread from the center of its action (a definite
bristle) equally in all directions. As he assumes that the decisive
agent is a catalyst, he thinks that this diffuses from its point of
origin in all directions, a process, then, that we might describe

200 PHYSIOLOGICAL GENETICS

also as a determination stream. Other cases of the same type
will li.' mentioned in the oexl chapter (see page 229).

In plants, we have already mentioned the determinative proc-
esses in Acctabularia (Haemmerling) dependent upon the flow
of a substance which may be called a determination stream.
Such a process in relation to definite genes would probably be
found if a genetic analysis of flower abnormalities were combined
with nnbryological experimentation. A not very clear hint that
such is the case may be derived from some remarks of Zimmer-
mann (1934) concerning hereditary traits in Anemone. Also,
mutations of the type of laciniate leaves might permit an analysis
in that direction.

9. THE PROBLEM OF PATTERN

In the preceding chapter, we were confronted with the problem
of pattern, the most important problem for both the geneticist and
the embryologist. Development is, of course, the orderly produc-
tion of pattern, and therefore, after all, genes must control
pattern. It is a formation of pattern when the axial organs in an
amphibian are determined; it is also formation of pattern when
the digits of a hand are laid down, as well as when two blastomeres
of different potency are separated. Most of our knowledge of
pattern formation comes from experimental embryology, which
is the science of analysis of pattern formation. We must now
try to find out whether genetics has furnished material that per-
mits an attack upon the problem of pattern in terms of gene
action. Before doing this, however, we ought to have definite
notions regarding this problem, as it emerges from the facts of
experimental embryology, according to our ways of interpreting
these facts.

A. Pattern in Development

Development, if considered apart from the component cells,
consists of a series of steps restricting the potencies of different
areas, which are thus consecutively subdivided into smaller and
smaller areas of more and more specialized determination.
For each part or organ or structure, then, sooner or later comes a
time whence its future is finally determined1 or, in the language

1 Harrison (1937) Wants the term final determination eliminated because
each determination is not final. But as it is final in normal development,

THE MUTATED GENE 201

of Driesch, whence its prospective potency is identical with its
prospective fate. The processes that lead to this type of pattern
formation are usually conceived as being of two types, self-
differentiation and dependent, or inductive, differentiation.
But it may be regarded as certain now that these two types of
developmental activity are, in fact, only two phases of the same
process. If we try to describe an inductive process in a most
general way, it would be one that leads to pattern-like restriction
of potencies in an area outside the area containing the inductive
agent. Induction of pattern by a process or product of a process
within the area to be subdivided would also be induction (see
page 202). It would then be an inductive action if a hormone
were sent to the area of induction or if a substance formed in the
inductor were to diffuse into the area to be induced or if a con-
tact between two areas, if established, were to produce an
unknown physical effect, followed by pattern formation.1

The best known inductive agency is the organizing center of
amphibians. We know that its active substance, at least for
the general inductive effect, is a relatively simple chemical com-
pound (though unanimity is not yet established regarding its
actual nature); therefore the inducing action must be conceived
as diffusion of this substance into another area. This process,
then, would be comparable to the determination stream already
mentioned, though not completely. Furthermore, we know that
there is a regional differentiation within the inductor area, which
has different inductive effects, e.g., a part that may induce only
head pattern. This shows, then, the presence of a pattern within
the inductive area, whatever it is. An inductor, then, is some-
thing rather general (not very specific) ; something that may show
a certain diversity within this generalized nature; something that
is moved to adjacent areas; something that makes the area upon
which it acts diversify, form a pattern, the details of which
depend upon the constitution of the inductor substance and upon
the condition of the area to be induced. Finally, the first set of
inductors may be followed by a second or secondary one within

controlled by genes, and also final in the average type of experimentation,
we feel entitled to use this conception.

1 The general literature on experimental embryology will not be quoted
in this section, except when special examples are used in connection with the
topic of this book.

202 PHYSIOLOGICAL GENETICS

the newly diversified areas, and so on. Induction, then, is ;i
process that provokes pattern formation but docs not explain it.

Self-differentiation means ili.it for the area in question pattern
formation has been ended. It will not be subdivided any more.
[f it occurs late in development, it will he the final outcome and
the end of the series of inductive processes. If it begins early in
development, it means that pattern formation has occurred
early. In these cases, it might have been preceded by induction,
but it might also be that induction had taken place so early in
development, even within the egg, that it would actually coincide
with pattern formation. This would mean that the inductor
is produced within the area upon which it acts as an evocator of
pattern. (Example: the induction of pattern in the egg as a
consequence of the bursting of the germinal vesicle.) Induction,
then, is one of the methods of initiating pattern formation. It
therefore does not explain pattern formation. This is an impor-
tant point, because the beauty of the experiments with the
"organizer" has led to a general belief that here the causative
principle of development was discovered. In fact, only an
evocator (Waddington) was found which in certain stages and in
certain cases acts like a fuse to set in motion the actual pattern-
forming process.

The following processes, then, have been found by experi-
mental embryology to combine in development, according to
our notion:

1. The ability of different embryonic areas to be broken up
into subareas according to a definite pattern; a process to be
followed later by a similar subdivision of these subareas.

2. The possibility that these pattern formations may be started
at different times in development of different species, either in the
whole embryo or within different parts of the same embryo.
Late patterning means equipotential development; early pat-
terning, mosaic development.

3. The breaking up of areas (or embryonic fields) into sub-
areas finally leads to a point where each area is so restricted in
potency that only one type of further differentiation is, as a rule,
possible. The area is determined.

4. The breaking up of one area into subareas, the process of
patterning occurs in orderly fashion both as to time and as to
localization in normal development. Regardless of all other

THE MUTATED GENE 203

control of development by heredity, it is first bound to control
this sequence of patterning.

5. The process of pattern formation may be evocated by
evocator substances diffusing into the area from an adjacent area
or center (organizer type) or by substances diffusing from larger
or smaller, already formed areas in definite directions. In the
latter case, it cannot be decided whether an evocator is actually
present or the observed process of diffusion, e.g., sea-urchin egg,
amounts to separation of the pattern-forming substances. Some-
thing, however, must initiate this process and might be called
an evocator, produced within the area upon which it acts, but
this need not necessarily be a substance; it might also be the
pH situation or an electric charge, etc.

6. The type of pattern formed depends upon the nature of the
evocator (example: head and tail organizer), the nature of the
area (example: newt evocator acting upon frog-mouth area),
and probably also upon the conditions of the whole developing
system (example: different success in transplant and explant.)
The problem of regulation has been barred from this general
review. This is how we conceive of the general picture of devel-
opment as derived from modern work. (The facts of experi-
mental embryology are assumed to be known to the reader.)
The action of the hereditary material, then, has to be linked with
these primary processes.

Before we begin to study the facts, some information should be
presented regarding the decisive process of patterning. Genes
or mutant genes that control development ought to be responsi-
ble for the initiation of such a process at definite times and
places by some type of evocation. What follows, the subdivision
of the field, may be a wholly automatic, i.e., physicochemical,
consequence of the situation created. The actual process of
pattern formation must therefore be of a rather generalized type;
as a matter of fact, it may be any type of process that leads to a
redistribution of substances within a given system (or eventually
of physical conditions like viscosity, charge, dispersion).

A considerable number of investigators have tried to formulate
definite ideas of this process. One point of view, couched in
physiological language, is Child's theory of gradients which,
applied to our problem, would mean that a gradient of intensity
of metabolic processes produces different conditions in a set

204 PHYSIOLOGICAL GENETICS

direction, which will be followed by morphogenetic differences.
The evocator then would determine the center of a metabolic

gradient, and the different levels of the gradient would determine
the different parts of a pattern.

Another viewpoint of a more general type is Lillie's theory of
segregation (1929). According to Lillie, embryonic segregation is
the process of origin of the diverse specific potencies that appear
in the organism in the course of the life history, which express
themselves later in tissues of specific structure and function.
The criterion of segregation is self-differentiation. Segregation
proceeds from the more general to the more special until
further segregation is closed. Thus far, this conception is
practically identical with all other similar conceptions. Lillie,
however, postulates that this segregation is dichotomous. It is
not clear whether he means only a dichotomy into parts with
restricted and nonrestricted determination or thinks that simul-
taneously only two areas may be formed, not a complete pattern
of areas with different restricted potencies. This dichotomy may
coincide with cell divisions in the case of cell lineage, and it may
also cover multicellular areas. Only a comparatively small
number of such segregations occur in ontogeny. Inductive
processes are those wdiich cause segregation to take place.
Segregation, however, does not separate organ-forming stuffs
but possibly only different proteins (which is the same, Author).
This theory, though couched in somewhat different terms, is
after all not very different from others here reported.

Another more generalized point of view is the theory of
embryonic fields in the slightly differing types worked out by
Gurwitsch, Weiss, and Guyenot (see Weiss's review, 1935).
They compare the area that is to be subdivided to a field of
force (without making special physical assumptions) that is
being subdivided by the evocation of new conditions of equi-
librium. (This is only a very generalized statement of these
views.) Another view has been developed by Goldschmidt
(1927c). He uses as an inorganic model the formation of
Liesegang rings in a colloidal solution, evoked by the infusion of
a definite substance. With a generalized expression he calls such
a diversification within a unit, produced by a chemical or physical
evocation, a process of stratification; this might work via all
known chemical or physical processes such as change of charge,

THE MUTATED GENE 205

processes of diffusion, separation, changes of viscosity, adsorp-
tion, chemical equilibrium, sedimentation, etc. In Liesegang-
ring formation, the nature of the solution, the nature of the
evocator substance, and the whole system (arrangement and
texture of filter paper, etc.) will determine what the pattern of
the rings will be. In development also, the nature of the evocator
and of the area to be patterned, together with the conditions of
the whole system, will decide the pattern within this area. Many
different processes of diversification within a given physico-
chemical system after the model of Liesegang rings are imaginable.
All have in common that an act of evocation (in the model,
putting a drop of a certain chemical to a certain point) auto-
matically produces a definite equilibrium arrangement of present
or newly formed different substances in the form of a pattern,
a process designated by the general term stratification. (A
point of view practically identical with ours has recently been
derived by von Ubisch (1936) from his work on sea urMrins and
also by Lindahl (1936), who actually established some chemical
patterns.)

Finally, Harrison's views (1937) may be mentioned, though he
has thus far pointed out only the general direction of his delibera-
tions. He thinks that change in the spacing of the atomic lattice
within the molecules of living matter may turn out to be relevant
for cytoplasmic differentiation.

The important thing that is at the basis of these formulations,
and that characterizes the developmental process which ought to
be linked directly with the action of the genes, is the automatic
physicochemical process of pattern ( = equilibrium) formation,
given evocation at definite time and place and the conditions of
the whole system and its parts.

B. Genes Controlling Developmental Processes

A discussion of the problems of development with the ultimate
purpose of understanding the action of the genes presupposes
that the decisive steps in development are controlled by gene
activity. Most biologists who have discussed the problem of
whether the promorphological features of the egg before fertiliza-
tion— early formation of pattern (organ-forming areas) — or
in early development are controlled by genes have come to the

206 PHYSIOLOGICAL GENETICS

conclusion that this is the case (Boveri, Wilson, Conklin, Morgan,
Goldschmidt, etc.). Though it would be difficult to contradict
such an assumption, the question ought to be answered whether
or not facts are known supporting such a self-evident assumption.
Actually, there are not many such facts. This is to be expected,
as only rare cases are imaginable in which the action of mutant
genes upon Buch early processes could be analyzed. But those
which are known justify such a generalization. There is the
work of Toyama (1909) and Tanaka (1924) on characters of
silkworm eggs. These are different colors of yolk, already exist-
ing in the unfertilized egg, and colors of the serosa, a membrane
formed by the blastoderm cells. These early embryonic
characters are determined by mutant genes. (For an inter-
pretation see the pigmentation hormone on page 184.) A
more direct relation to the pattern problem is provided by the
facts regarding right- and left-hand coiling in molluscs. It has
been knoVn for a long time that this may be a modification in
some cases and a hereditary trait in others. The genetic analysis
of such cases by Boycott et al. (1930), Sturtevant (1923), and
Crampton (1924) showed a Mendelian behavior writh the addition
that Fi behaved like the mother, F2 like Fi in the classic case, and
F3 like F2. This means, as first shown by Toyama, that a
character of the egg is involved that has already been deter-
mined before fertilization, so that the influence of the sperm
will be felt only in the eggs of the next generation. The embryo-
logical work of Kofoid (1895) and Conklin (1903) had actually
shown that the direction of coiling is already determined when
segmentation begins; i.e., it is laid down in a promorphological
structure of the egg. Here, then, certainly, one of the primary
processes of pattern formation is controlled by mutant genes.
Another example of this type is furnished by Tanaka's work,
already mentioned. Among the characters of early stages that
wen1 studied there was one type of pigmentation of the blastoderm
which was inherited in the way just described (wrongly so-called
maternal inheritance), which proves determination before
fertilization. Blastoderm formation is a form of pattern forma-
tion of a special type in the insect egg, and the pigment thus
serves as a marker of the separation (stratification) of one
embryonic area controlled by mutant genes. A similar case has
also been reported by Sexton and Pantin (1927) for larval lipo-
chrome colors in Gammarus.

THE MUTATED GENE 207

There are several cases involving this so-called maternal
inheritance, in which it is difficult to form an idea of the type of
prefertilization influence of genes upon cytoplasmic structure.
In some of these cases, it is not even completely certain whether
or not the interpretation is correct. Thus, Morgan (1912,
1915) claimed a prematuration effect to explain irregularities in
the inheritance of the gene rudimentary; Lynch (1919) applied
the idea to an explanation of sterility in rudimentary and fused
(Drosophila); Redfield (1926) claimed maternal inheritance for
a sex-linked lethal in Drosophila which acts only when the mother
is homozygous for a second chromosome gene; Gabritschevsky
and Bridges (1928) found a maternal effect upon the eggs of
homozygous females for an enhancer gene of giant in Drosophila.

Rather remarkable are such cases of maternal inheritance in
which the effect of genes acting upon the organization of the egg
cytoplasm prior to fertilization becomes visible only in later
developing parts of the body. Dobzhansky (1935) found such
an effect in reciprocal crosses of two races of Drosophila pseudo-
obscura, controlling the size of testes. An autosomal gene acting
on the egg plasm is responsible. In this same form, Dobzhansky
and Sturtevant (1935) described a number of further hereditary
characters with maternal inheritance, viz., viability, sex ratio,
rate of development. The latter characters may be understood
more easily as controlled by a general plasmatic condition induced
by genes. There is also a statement by Warren (1924) that egg
size in Drosophila is controlled thus. It is more difficult, how-
ever, to understand the findings of Timofeeff (1935) that a gene
controlling the pattern of bristles in the thorax of D. funebris
(polychaeta gene) shows maternal inheritance. A predetermina-
tion of a pattern of thoracical epidermis in the cytoplasm of the
egg is more than could be expected even in cases of develop-
ment with very early determination (self-differentiation).

There are many facts showing that details of later stages of
development are influenced by mutant genes. The whole body
of work upon Mendelizing characters of larvae of insects (silk-
worm, Lym,antria dispar, and L. monacha by Toyama, Tanaka,
Goldschmidt, etc.) could be mentioned. There are also larval
characters in Drosophila controlled by genes (Chubby, Dobzhan-
sky and Duncan, 1933). In plants, all the endosperm and coty-
ledon characters belong in this category, also characters of the
protonema in mosses; moreover, the influence of genes upon rate

208 PHYSIOLOGICAL GENETICS

of segmentation (Castle and Gregory) ought to be mentioned,

and the whole series of let hals which act upon early development.

Nearly related are the cases in which mutant genes influence
directly early pattern formation in vertebrates, though the
exact type of influence is not always know'n. These are the cases
of monstrosities in vertebrates which we have already reported
(see page 46), some of which demonstrate a genie action at a very
early time of development, probably upon the organizer directly
(see Lehmann, page 50). To this category belong also all such
lethal genes as are known to kill the embryo in very early stages,
probably by upsetting primary pattern formation, e.g., the cases
analyzed in Drosophila by Redfield (1926).

Another group of most important facts has been derived from
the study of a phenomenon called homocosis. We have described
on page 36 the development of the mutant aristapedia in
Drosophila in which the last segment of the antenna is trans-
formed into a tarsus. We saw that this differentiation of a
homologous organ takes place at the time of leg differentiation
which occurs prior to the time of differentiation of the antenna.
Here, then, a mutant gene changes an embryological process
by shifting its initiation to a different point in time. There are
similar cases in Drosophila which, however, are not so clear as this
one. There are the mutations bithorax and tetraptera (see
Astauroff, 1929) in which the metathorax assumes a number of
characters of the mesothorax, including the formation of wings
instead of halteres. There cannot be any doubt that the explana-
tion wrill be of the same type as for aristapedia. There is also
the mutant proboscipedia (Bridges and Dobzhansky, 1933).
Here the mouth parts undergo a number of variable changes
which bring their structure near to the type of biting organs of
other insects. The modified lobes may show resemblance to
the labrum, maxillary palpi, antennae, and tarsi, the homology
of all these organs being apparent. Again, the same type of
interpretation applies as in the case of aristapedia. These facts
are of greatest importance in linking gene action with develop-
mental processes. They point out that the mutant gene in
question does not control a localized action of a special type. It
controls a general, quantitative process which leads to a special
result, because it shifts a process of pattern formation into a
phase in which it ought not to be. To be more specific, at the

> THE MIT ATED GENE 209

time when the pattern of the tarsus in the leg disks is laid down
by a process of rhythmical subdivision of the distal end of the
Anlage, something, an evocator, is present in the germ that
diffuses into the Anlagen of the segmental appendages to produce
this result. Disks in the proper stage of development will
react to this induction by tarsus formation. Normally, the
antennal disk is far behind in differentiation at this time and is not
influenced by the evocator. The mutant gone, which speeds up
antennal differentiation, however, makes the antennal disk
mature simultaneously with the leg disks; and the evocator
substance, which "orders" the formation of tarsus segments,
therefore also acts in this disk. Here we have a case where a
simple shift in the time element of gene action results auto-
matically in a complicated morphogenetic change. Such facts
show a system of genie action at work which we have discussed
in a former chapter as the system of reaction velocities in tune.
Here we may finally draw attention to the great evolutionary
significance of such facts in connection with the problem of single
large steps, the problem of the "hopeful monsters" as Gold-
schmidt (1933c?) has called it, which is much more serious than
this terminology would indicate.

C. Genes and Pattern

The facts just described justify the conclusion that the process
of embryonic stratification is controlled by the action of the genes.
To gain further insight we must, however, turn to more easily
accessible forms of pattern, which may serve as models from
which generalized conclusions upon all kinds of pattern may be
drawn.

Different types of pattern are available for such studies.
As the most elementary type may be considered the simple
growth pattern, i.e., the pattern-like typical differences in growth
in the three dimensions which lead to the establishment of a
definite form of all developmental units. Much work has been
done in this line to establish the general laws of differential
growth, to mention only D'Arcy Thompson (1916) and J. S.
Huxley (1932). The insight that these authors gained has
to be linked with specific genie action which might set in motion
the pattern of differential growth.

210 PHYSIOLOGICAL GENETICS

It is only a short step from a growth pattern to a pattern of an
area type, as the work Oil certain patterns in insects has shown.
Long ago. GoldschmicH (19206, d) pointed out that the insect
wing and especially the wing of Lepidoptera might serve as a
model because rather complicated patterns of form, structure,
and color here arc found in a more or less bidimensional arrange-
ment, and because this pattern is accessible to genetical as well
as embryologies.! experimentation. The first steps in this
direction were published in 1920c/; more material was added later;
and a comprehensive theoretical treatment was given in 1927c.
Recently this problem has been attacked mainly by Henke,
Kuehn, and their students, and the body of facts now available
furnishes considerable information.

These two main types of pattern formation have some addi-
tional special aspects. There are patterns of primary importance
for the progress of development, e.g., the pattern that is laid down
in the formation of an insect wing with the proper arrangement
of veins, etc. But there are also secondary patterns which are
superimposed upon the primary patterns late in development,
e.g., melanism, spread of dark pigment, upon an otherwise
completely patterned wing. There is, moreover, the problem
of symmetrical arrangement of the patterns on both sides of the
body, a problem that might furnish decisive information upon
the pattern problem in general. The following chapters will
deal with these and other aspects of the pattern problem as
related to gene action.

1. Pattern of Growth and Form. If we refrain from a detailed
discussion of differential or heterogonic growth (which is found
in the book of Huxley, 1932), we may state in a few words the
points that are essential from the standpoint of physiological
genetics. Very different and specific forms and shapes in two or
three dimensions may be produced if the process of growth in
different directions occurs according to a definite rule, containing
variables the values of which produce a definite result. Huxley
found an exponential formula that accounts for all cases of
differential growth and contains two decisive constants. The
value of these constants determines the numerical system that
the differential growth will follow in each case. Whatever the
specific merits of this formula may be (see discussion by Hersh
and Feldman, 1936), it furnishes, if considered in a general way,

THE MUTATED GENE 211

a clue to the possibility of linking gene action with the produc-
tion of growth patterns: any gene that controls quantitatively the
rate of any one process, which is represented by the constants of
the formula, automatically produces a definite result in terms of
form. Form, then, would also be of the type of automatic
pattern formation: something connected with growth produced
at a definite time and place and in a definite quantity sets the pace
for the working of differential growth by fixing the numerical
value of the constants in the formula; it acts in this sense also as
an evocator of pattern.

This important conclusion had already been derived by Gold-
schmidt (19206) from D'Arcy Thompson's (1917) classic work
on growth and form (including also the phylogenetic significance,
which recently has been emphasized again by Huxley, 1932,
and tested in an admirable way by Hersh, 1934a). Goldschmidt
wrote :

Let us suppose that we succeeded in reducing all processes of morpho-
logical differentiation to simple physical and mathematical conditions,
to surface tension and to all the geometrical consequences of differential
growth, etc., the problem of heredity remains the same: to find why, at
a given moment in given rhythm and localization, the specific chemical
and physical situation is produced, the consequence of which may be
stated as the mathematical law of development. . . . [Here follows a
statement of the theory of rates of reactions in time.] . . . This quanti-
tative conception makes it possible to refer extraordinary differences in
the end result, i.e., evolution within considerable limits, to very small
causative processes. One may find in Thompson's brilliant book how
the production of all forms of shells may be reduced to small changes in
the values of the terms of the underlying equations for the actually found
curves ; and therefore how then a series of morphological differences may
be shown to be produced by changes in differential growth within simple
mathematical laws. Remembering our proof (in the work on inter-
sexuality) that very specific differential growth may be started by the
production of specific hormones [here in the sense of determining stuffs,
evocators] at a definite time, one might realize the number of evolu-
tionary processes which might be caused by small changes in the under-
lying genes followed automatically by numerous shifts in the interplay
of properly timed coordinations.

The genetic facts that permit the derivation of such views are
only beginning to be studied quantitatively. In earlier chapters,
we have mentioned a number of facts that have to do with

212 PHYSIOLOGICAL GENETICS

determination of growth. Before analyzing differential growth,
we may review shortly the pertinent genetic and other facts
concerning growth in general. In animals, as well as in plants,
the size of an organ (or the whole body) depends upon two proc-
esses, cell division and increase in cell size. Usually these follow
each other in individual development ; i.e., at a certain time, cell
division ceases, and growth of cells begins. There are many
examples, one of them found in the butterfly wing where epidermal
cell division ceases when the formation of scales with corre-
sponding cell growth begins. In the Drosophila wing, we have
seen that cell growth begins after the structure of the wing is
completed. The same two processes are found in the develop-
ment of a bird's feather or in the growth of a fruit. The most
extreme case is found in such cell-constant animals as the
nematodes (Goldschmidt, Martini), the Rotatoria (Martini), and
the Acanthocephala (Van Cleave) in which a limited number of
cells composes all organs, cells that might grow to an immense
size without further division. Genetically, the two consecutive
types of growth may be controlled by different genes. It is
known that in Drosophila the wing-mutant Miniature prevents
only the cell growth following a normal development during the
phase of cell division and differentiation (Dobzhansky, 1929;
Goldschmidt, 1935c, 1937). The wing mutant Expanded (large
broadened wing), however, is produced by a different amount of
growth in the period of cell division (Goldschmidt, 1937). It
is possible to combine both genes in one individual, with the
result that a broad wing is formed in the first phase which, how-
ever, does not grow in the second (Csik, 1934a). These facts
show that the size produced by growth may be controlled
by genes affecting at least these two processes of growth, maybe
differentially. But differences in size may also be influenced by
other genes acting on growth in general, irrespective of the two
phases. It is therefore to be expected that in different cases
different relations might be found. To mention some: Hough-
taling (1935) has studied fruit size in tomatoes which, according
to Lindstrom (1928), is controlled by a number of genes with
different effect. He found that in the ovary of the large forms
more cells are present and that, in addition, larger cells are found
in the larger fruits; i.e., both processes of growth are changed by
mutant genes. The first, cell division, is ended about flowering

THE MUTATED GENE 213

time and has then already produced size differences in the
different stocks. Then growth of the individual cells begins and
also lasts relatively longer in the larger races, thus producing
larger cells. In this case, then, both phases of growth are deter-
mined together at an early stage (as opposed to the quoted case
in Drosophila). A similar result was also derived by Sinnott,
Houghtaling, and Blakeslee (1934) for polyploids and trisomies
of Datura, where it is concluded that some genes control increase
in cell size, others increase the rate of cell division. Somewhat
different is the case of inherited size differences in Lymantria
dispar (Goldschmidt, 1932d, 1933a). Here the growth by cell
division starts with eggs of the same size in races of different
genetic size, the differences of which are controlled by a number
of genes. As is indicated by the growth curves of the larval
stages, cell division is faster from the beginning in the larger
races (with the addition of sexual differences between the large
females and small males). This agrees with the findings of
Castle and Gregory (1929) in rabbits of different hereditary
size. But in Lymantria there is also an additional growth by
increase in cell size, as it was proved that size of spermatocytes
and of wing scales — two types of cells with considerable post-
division growth— increases in proportion to inherited body size.
But the same increase of size of these cells is also found in larger
plus variants of a small race. The nonhereditary factors, then,
act in the same direction upon cell size. This might mean —
though not necessarily — that the second phase of growth is
dependent upon actual size reached at time of pupation which in
all cases will be followed by a definite percentage cell growth.
These few examples show that growth of a body or organ may be
determined by genie control of one primary process of rate of cell
division, possibly by the production of growth hormones at
definite times and in definite quantities; further, by control of
the secondary process of increase in cell size (in plants known to
be under auxin control); further, by control of the relative times
of beginning and ending of both processes and also of their
relative intensity. The integrated effect of such of these ele-
ments as may be present in the individual case is the measured
linear growth.

Now to the problem of this chapter — the pattern of form by
differential growth and its eventual explanation by a single or a

214 PHYSIOLOGICAL GENETICS

few genie actions, controlling one of the just-mentioned variables

within the parts that grow at a differentia] rate from others,
thus producing a pattern of form. We have already met with
cases which might be described in this way, when we studied the
production of hereditary abnormalities by changes in definite
areas of early embryonic stages: the rimipless fowl, the mice with
deformed tails, etc. Another example, which we mentioned there
briefly, the creeper fowl, shows, however, that here a different
type of production of form pattern is involved. According to
Landauer and Dunn (1930) and Landauer (1931, 1934), the
creeper fowl, a short-legged fowl, is the result of a dominant
mutation which is practically lethal when homozygous. The
limbs here were shown to be smaller even in the earliest stage of
their visibility, and all their further growth occurs at the normal
rate. The special pattern of Short-leggedness is then not
produced by a change in the variables of growth of the limb
but by a change acting only upon the early embryonic Anlage
of the limb — a slowing up of differentiation in the early deter-
mination field of the limb. (But there is also a smaller effect
upon growth in general.) The various bones are not equally
affected. The more distal bones in the limbs wrhich are deter-
mined later are more affected; the proximal ones, wdrich have
at least partly been determined before the retardation within
the Anlage set in, are less affected. The fact that the longer
bones are more affected indicates that tissues with the larger
growth intensity are more affected by the retarding action in
the Anlage. In agreement with this interpretation is the fact
that the homozygous embryos which die at an earlier stage show
a large retardation in growth during the first days of incubation,
which most strongly affects those organs that are found at this
time in their most active stage of growth. Landauer, there-
fore, concludes that a general retardation of growth in early
stages, wrhich become localized by the more ready response of
parts in most active growth, is responsible for all phenomena.
If this interpretation is correct, the case does not actually belong
to the group of patterning by differential growth, because the
pattern is produced by a change in a general rate process, which
automatically affects more strongly such parts as are in active
growth. Probably the same explanation applies also to the
other cases of a similar type (reported on page 47ff.), according

THE MUTATED GENE 215

to the view taken by Sinnott and Dunn (1935), who cite an addi-
tional case from their laboratory (kinked tail in mice, after
Kamenoff (1934)).

But information is also available for animals that suggests
that the genie control of differential growth is one of the
means of pattern production by simple effects of genes upon one
process with the consequence of an automatic patterning.
Among the phenocopies that Goldschmidt (1929a, 1935a)
produced in Drosophila by heat shocks was one that consisted of
a change of the wing shape into a lancet form; all transitions
from normal to lancet shape could be produced, apparently as a
result of the establishment of a new differential growth in
length and breadth of the wing in favor of length. In this case,
only the second phase of growth, by enlargement of cells, has
been involved, and mutants of this type are not yet known (?).
But the simple quantitative increase of the heat effect in the
series of types suggests a rather simple change of rate of some-
thing as the basis for the production of the different shape.

Another case, in which the genetic side is also known, is the
case of the Bar eye in Drosophila according to Hersh (1928).
It is known that the eye of Drosophila (and other Diptera) is
— visibly or invisibly — subdivided into dorsal and ventral lobes.
In the Bar-eye mutants, these lobes become clearly visible.
Hersh found that the temperature effects upon the eye are differ-
ent in both lobes, the ventral lobe decreasing at a faster rate with
a rise in temperature. A statistical study of the data for different
genetic compositions led to the conclusion that formation of
facets (or destruction of facet-forming substance, Author) occurs
according to Huxley's equation for heterogonic growth. As the
ommatidia are differentiated in the optic disk primarily by an
induction from the brain (see page 34), the growth pattern
(ventral and dorsal lobe) may be the consequence of a gradient
(determination-stream, Author) or of a primary inductive differ-
ence. However this may be, Hersh concludes that the genes
of the Bar series produce their effects by altering the distribution
of growth in the developing zygote. As he finds that Ultrabar
larvae grow faster, it may be assumed that the facet-forming
reaction will have a different time relation to the general curve
of growth. "The manner of action of genes which affect the
size of the eye, as determined by the number of ommatidia, would

216 PHYSIOLOGICAL GENETICS

seem to reduce itself to a question of the distribution of growth
in the zygote." (It is of interest to note that exactly the same
conclusion, though couched in somewhat different language, had
been derived by Goldschmidt, 1927c from a general theoretical
analysis of the case. )

In a later paper, Hersh (1934c) could specify these statements
by calculations based on Driver's data, and this new statement
gives exactly the type of information for which we are looking
at this point of our discussion. The relative growth function,
according to Huxley, is y = bxk which means that if y increases
by a certain percentage, the variable x increases by another cer-
tain fixed percentage. The ratio of the percentage increases of
x and y is the value of k, the coefficient of growth partition (b is
another constant relating to initial size of the organ). The data
suggest now that k is constant for a given temperature: at 28° the
larval length x increases by about 1 per cent with an increase
of the facet number of 5 to 7 per cent. But it is different at
different temperatures. The value of k would then be con-
trolled, ceteris paribus, by the Bar gene, and the temperature
effect is of the same type as discussed previously, including the
possibility of a phenocopic effect. Hersh further is inclined to
draw conclusions upon the facet-forming process involved (which
belong more to the discussions in a former chapter). From
the sigmoid curve which he obtains for the reactions involved
(facet-forming reaction) he concludes that it is not a sudden
arrest of the facet-forming process in the members of the Bar
series in relation to the general growth processes (which had been
assumed in Goldschmidt's theoretical analysis of 1927c) but a
process that approaches its termination asymptotically. After
Goldschmidt's recent work on the vestigial alleles, this result may
point in the direction of the possibility that in the Bar case, also,
destruction of facet-forming substance at an early stage is
involved, especially as Chen has shown that the Anlage of the
Bar eye is in the beginning no smaller than that of Full eye (see
also page 76, the work of Margolis).

Important contributions to the problem under discussion have
finally been made by Sinnott and his students in plants. Though
the growth of plants is frequently not a closed system as in
animals, there are many cases where it follows the same rules and
where also the differential growth formula applies (Pearsall,

THE MUTATED GENE

217

1927). Such cases, e.g., fruit size, therefore permit a similar
treatment, and they are pertinent to our problem when hereditary
differential (heterogonic) growth is involved in the form of
hereditary control of shape. The problem is then obviously the
same as in the Drosophila wing (see p. 212) though somewhat
complicated by the tridimensional aspect. Nearer to the
Drosophila wing are the shapes of leaves which, as is well known,

Fig. 39. — Diagrammatic representation of one half of a spread feather germ
(collar) of a regenerating feather. The germ has been split along the ventral
superficial axis (left) and also in the center of the shaft (right). Barbs and shaft
are drawn at right angles to the base of the germ. Cv Cd, locus of primary
growth by cell division which is extended to the level h0; Pv Pj, a locus of simul-
taneous pigmentation. Ridges join the shaft at u. (From Fraps and Juhn,
1936, Phys. Zool. 9, Fig. 2.)

are controlled by Mendelian genes in innumerable instances.
We have pointed out that size and shape of fruit are controlled by
Mendelian genes, sometimes by a single one (tomato, Lindstrom,
1928). In Cucurbita pepo (Sinnott, 1927; Sinnott and Ham-
mond, 1930), two independent allelomorphs with cumulative
effect (cf. the Drosophila case, page 212) control the disklike or
spherical shape of the summer-squash fruit. In this case,
Sinnott (1935a) made it clear that shape is not a physiological
consequence of size but controlled by independent genes, as is

218

PHYSIOLOGICAL GENETICS

proved by statistical study of the hybrids between shape races.
(As a matter of fact, Lindstrom (1928) had already distinguished
between genes for size and shape in the tomato.) A study of
development showed, then, that the underlying genes actually
control the value of k in the heterogonic growth formula, the
coefficient of growth partition. This is well demonstrated in
Kaiser's (1935) work on Capsicum, where after flowering time,

fruit shape changes according to the
heterogonic growth formula with a
definite value of k. A similar case
is found in the tomato, where, how-
ever, according to Houghtaling
(1935), the heterogonic growth is
confined to the early development
before flowering. In the squash, this
period occurs still earlier (Sinnott and
Kaiser, 1934).

At this point, the work of F. R.
Fig. 40.— Diagrammatic trans- Lillie and his school on the bird's

verse section of the base of the feather ought to be mentioned,

feather germ. D, dorsal limit; V, .

ventral limit. Initially barbs are though it is as much Concerned with

laid down from D to i; later, at pigment pattern as with growth pat-

v. They are carried by tangential r ° r

growth from v to u. s, shaft , pri- tern (Lillie and Juhn, 1932; Juhn,

mordium increasing in the direc- Faulkner, and GustafsOIl, 1931; Juhll
tion of the arrows. (P rom Fraps .

and Juhn, 1936, Phys. Zooi. 9, Fig. and Fraps, 1934). Also, in feather
8-) development, a growth by cell divi-

sion is followed by enlargement of cells. The growth by cell
division takes place, however, not after the fashion of a typical
growing animal organ but more after the fashion of growth from
a cambium zone in plants. This zone is found in the circular
collar of the feather follicle, from which at one dorsal point the
shaft is growing out while the outgrowing barbs are lined up
around the collar. The growing barbs are constantly shifted by a
circular growth process toward the shaft and unite with it, when
they have reached a certain length. The new supply of bar!)
material is simultaneously pushed up from the ventral side of the
collar ridge. Figure 40 of a transverse section through the collar
and Fig. 39 showing one-half of the feather germ spread out will
help in understanding these relations. Thus, loci of simul-
taneous age of the barbs are situated at different points on the

THE MUTATED GENE 219

individual barbs, viz., nearer the tip in the barbs which grow out
more vent rally as compared with those growing out toward the
shaft. Lines of equal age are therefore parallel to the base of
the collar and stretch transversely across the feather. Now,
injections of thyroxin produce pigmentation in an otherwise
light-colored feather. Lillie and Juhn showed that with smaller
concentrations, pigment appears only near the shaft but stretches
across the feather with higher concentrations. Thus, there is
wTithin the feather germ a different threshold of reaction to
hormones, lower near the shaft and increasing toward the tips
of the barbs. These tips have, however, a shorter period of
latency and react therefore faster to the proper concentration of
hormone. By using these reactivities and injecting different
concentrations of hormones at different intervals, many beautiful
types of pigment pattern may be produced. Lillie and Juhn
believed that the decisive difference in regard to the threshold for
response to the hormone was the different rate of growth of the
barbs at different points of the collar, which would have reduced a
part of the problem to the problem of differential growth. But the
recent mathematical analysis of Fraps and Juhn (1936a, 6; Juhn
and Fraps, 1936) has demonstrated that differential growth is not
involved in the case. It is very difficult to decide whether or not
these results throw light upon the formation of feather patterns
controlled by genes. Of course, where purely hormonic differ-
ences are involved, as in the patterns distinguishing both sexes,
which can be shifted at will in regenerating (or transplanted,
Danforth), feathers by adding the proper sex hormone, hormonic
action, together with thresholds and some yet unknown rhyth-
mical process may account for the pattern. This is different.
however, for genetically controlled patterns like the Plymouth
Rock barring of feathers, which does not respond to sex hormones
(see Danforth's, 1929, transplantations). Montalenti (1934) has
tried to attack this problem. He found that substances in the
blood cannot be regarded as responsible for the rhythmical
pattern. But there is one interesting relation: feathers in
different body regions have a different velocity of growth, and the
breadth of a double bar varies proportionally. If we think of the
facts regarding the velocity of differentiation of wing scales in
butterflies as related to the pattern formation (see page 241), we
might conclude that a similar process might take place in the

220 PHYSIOLOGICAL UENETKS

feather germ. Either a rhythmic production of a growth sub-
stance produces in the cambium zone of the feather germ an
alternation of quicker and slower cell division with the same
result a* in mollis, viz., that the cells that enter their growth
phase later are the ones that can produce the pigment; or, within
the cambium, a simultaneous "stratification" takes place very
early, resembling the Liesegang ring model (rings in test tubes),
and determines which cells may form pigment. In spite of much
interesting work, the essential problem — the pattern-forming
action of the gene — has hardly been attacked in this case.

Not much more is to be said about the related problem of the
agouti hair in mammals in which alternate zones of different
pigmentation exist. Dry (1928), who studied this case, ascribed
this pattern to an alternation in the hair follicle, conditioning the
quantitative aspect of a pigment-forming process in the following
sequence: high = black; intermediate = yellow; low = "low-
black. " No information is obtained concerning the causation
of the rhythm.

In concluding this section, we may finally refer to work
relevant to the problems under discussion but undertaken more
from the standpoint of experimental embryology than from a
genetical viewpoint. Embryologists frequently have made
transplantations of organs between animals of different hereditary
size (Harrison, 1924, 1929; Twitty, literature, 1934); and, as a
rule, the transplants kept up their own hereditary growth rate,
which is to be expected if the determination occurs rather early
(see above). But there is also an influence of the host, i.e.,
the environment visible, which acts upon the transplant by
functional, mechanical, or nutritive means. More difficult to
explain is the influence that different parts of the organ exercise
upon each other: a lens from a larger species working with an
eyecup of a smaller species gives a compromise result, each part
adapting itself to the different growth rate of the other (Harrison).
This belongs, of course, to the difficult subject of regulation. The
fact that such regulations usually require contact of the parts
indicates at least that principles not different from those involved
in ordinary induction may be at work. But there is also a
regulation of a different type. If older organs are transplanted
into a younger host, or younger ones into an older host, both are
influenced so that normal growth relations ultimately result

THE MUTATED GENE 221

(Twitty and Elliott, 1934). Twitty explains this regulation
by the assumption that the assimilative capacity of younger,
faster growing cells is larger than that of older ones, which would
mean that they will take out of the available nutritive material
a larger share, and vice versa. If this is true, it would follow that
the constant k of the heterogonic growth formula controls the
avidity of a group of cells to apportion and assimilate a share
of the total available food. Regulation then would mean
formation of a certain equilibrium in the apportioning capacities
of these cells. It is clear that such results do not change the
picture derived from the work in physiological genetics but that
they add to it a new intermediate step between gene and char-
acter: the value of k, controlled by gene action, becomes the
ability of cells to secure a definite share of available food.

Obviously, this is only one of the possibilities. It is generally
known that hormones play a decisive role in growth, e.g., the
molting hormones of insects (Buddenbrock, Wigglesworth), the
thyroid and thymus in vertebrates. No doubt, heterogonic
growth may also be controlled by hormones, as all the facts
concerning thyroid and metamorphosis, thymus and precocious
differentiation, sex hormones and growth of secondary sex
characters prove. There can be no doubt that among the prod-
ucts of genie action that control the growth pattern, hormones
of all the types discussed on page 181 play an important role,
including also all types of evocator substances.

2. Symmetry. — Wherever pattern is involved, growth pattern
or otherwise, the problem of symmetry will come in sooner or
later. Usually, the patterns on both sides of the body are alike,
but sometimes they are not. Such an asymmetry may be due
to simple fluctuations of the same process occurring on both
sides in development. It may also be of a typical hereditary
type, which may range from an inclination to asymmetrical
distribution to such extreme asymmetries as in the claws of
fiddler crabs or other Crustacea. It is known that the problem
of symmetry looms large in experimental zoology. There is the
problem of the primary symmetry in eggs, the formation of the
axes of the embryo which is studied wherever experimentation is
made on the early stages of development. There is the problem
of handedness as evidenced in innumerable aspects of form and
function (see Ludwig's monograph, 1932) . There is the statistical

222 PHYSIOLOGICAL GENETICS

study begun by Pearson and Duncker of right-left correlations to
determine their eventual genetic causation. There is the large
body of facts relating to symmetry and asymmetry in identical
twins (Newman, Bonnevie, Wilder, etc.). The innumerable
facts, both statistical and experimental, are relevant to the
present discussion only in so far as they permit linking the action
of the gene with the process of pattern formation.

Let us begin with certain facts that have been found in the
cases that we have already discussed. In the studies upon the
scalloping effect in the vestigial wing of Drosophila and similar
cases, as well as in many cases of bristle abnormalities in Dro-
sophila, the following fact was found by all observers. When
the very first steps of scalloping are produced either genetically
or by external agencies, i.e., the absence of a few hairs at the wing
tip up to a small nick, this happens almost invariably only in one
wing. This seems to be a rather universal phenomenon, when-
ever a genie effect is not strictly symmetrical. It may be
demonstrated for the unsymmetrical types of the mutants of
Drosophila, wherever the very low members of a series are
known, e.g., scalloped wrings of the beaded and beadex type, wing
venation of the plexus, delta, extra-vein type, eye defects of the
kidney type. It may be demonstrated also for the different
phenocopies. Another typical example is the mosaic-like struc-
ture of the wings of male intersexes in Lymantria, which are
shown in Fig. 35.

If we remember now the explanation given for the produc-
tion of scalloping, the first beginning of scalloping would be
produced by a small insufficiency of some growth' substance or,
reciprocally, by a small amount of lytic substance above the
threshold, acting late in wing development. If this were a
localized action of the mutant gene, both wings ought to react
simultaneously, except for a local variation in thresholds or the
like. If, however, the substance in question enters the wing
from its base, spreading through the wing, and if this substance
is produced in common for both wings, the amount distributed
to each wing is liable to considerable variation. For example, if
95 per cent growth substance is available instead of 100 per cent,
one wing may receive 46; the other, 49 per cent. If the threshold
for scalloping effect is 48 per cent, only one wing will be scalloped.
(In case of interpretation with lytic stuff: five units above the

THE M ( TA TED GENE 223

threshold present ; distribution, 3:2; active minimum, 2.5.) With
further increase of the effect, i.e., still more lowering of the avail-
able percentage, the probability of asymmetric distribution
with one wing above the threshold will decrease rapidly, and
absence of the character on one side will become very rare,
though asymmetrical expression on the two sides will continue.

For a general understanding of the process it is rather unim-
portant whether or not the foregoing interpretation is literally
true. The decisive point is that such conditions as quantities of
a growth-promoting or retarding substance, thresholds of its
action, time and place of its liberation, path of its spreading will
provide a system in which a definite constellation will auto-
matically lead to an asymmetrical result. Any gene controlling
the master process, which in a given constellation wrill enforce
asymmetric procedure, has, then, an asymmetric action. If
this explanation is correct, it follows that in cases such as those
just described, other genes (modifiers) may impress symmetry
upon the same development by shifting one of the integrating
processes, e.g., the value of the threshold or the time of onset
of the spreading of the substance. This is actually the case.
In the vestigial-wing series, the intermediate phenotypes are
rather asymmetrical on both wings. But it was possible to
select a symmetrical type from a few rare more symmetrical
specimens. 1

A special study of such a case has been made by TimofeefT
(1934). The character, already discussed on page 71, is the
formation of cross veins in the wing of Drosophila funebris. In
the case of the scalloping of the vestigial wing, and of the inter-
sexuality in Lymantria, there was an asymmetry for both wings,
but no large independent variation, because the average amount
of effect (scalloping, pigmentation) was about constant for a
given genetic constitution. In the D. funebris case, Timofeeff
thinks that variation was independent in both wings, and
asymmetry therefore a statistical result of two independent
variables. By selection it was possible to establish lines with
a larger right-left correlation, more symmetry. There are, then,
genes that may control the phenomenon, just as in the case
of the vestigial wings. Timofeeff remarks that one might think
of humorally controlled processes and points to the fact that in

1 Unpublished work.

224 PHYSIOLOGICAL GENETICS

Drosophila mutants mosl colorations are symmetrica] and
changes of structure of organs asymmetrical. He thus comes
near to finding the explanation which was given above for the

vestigial case ami which is also to be found more or less vaguely
expressed in other work on the problem.

A similar case occurs also in the work of Boycott et al. (1930)
on Lymnaeus to which we shall return immediately and which
belongs otherwise to a different type. Here a left-handed line of
shells exists, differing by a mutant gene from the normal right-
handed one. In these lines, right-handed forms also occur which
are not hereditary, although their percentage of occurrence is
hereditary. Whether this is caused by modifiers or by different
alleles for left-handedness, in some way an embryonic (here
cellular) situation near a threshold line must be involved, which
might be transmitted or not by some individuals according to
genetic conditions controlling a process involved in the threshold.
If it should be proved that such lines are actually based upon
multiple alleles, the interpretation would have to be parallel to
the one for the vestigial case.

We started our discussion of the problem of asymmetry with a
case that seemed to be instructive because the asymmetry con-
sisted in existence or nonexistence of the effect on both sides only
in the first steps of a quantitative series. However, the most
frequently studied types of asymmetry are different in various
aspects. We shall now consider three such types.

1. Constant asymmetry of an alternative type. For example,
a shell is hereditarily coiled either in a left or in a right-handed
spiral. (For examples of this and other types and their details,
see Ludwig's monograph and his recent review, 1936.) Here we
have the well-analyzed case of Limnaeus (Boycott and Diver,
Boycott et al.) in which a mutant gene changes the direction of
coiling (see page 206), and parts of which we have already
considered. Occasionally, however, a nonheritable variation
to the other type occurs (Lang, 1908; Boycott et al.), the fre-
quency of which might be hereditary (see p. 206).

In this case, the cytological basis of the asymmetry is known
(see page 206). In other similar cases, we must assume a parallel
process, i.e., the presence of a strictly alternative situation at
some point of embryonic determination where a determining
material for the formation of an organ or for controlling a thresh-

THE MUTATED GENE 225

old may go only to the left or to the right. (Ludwig formulates
this situation thus: the cells right and left have both potencies of
development, and a gene-controlled reaction decides that the
higher quantity of a decisive substance appears on one side.)
In cases in which one side only is favored by heredity, there must
be present some additional mechanism comparable to the one in
molluscs which directs a flow or orientation only toward one side.
Other cases, containing at least some genetic information, are
as follows. Beliajeff (1931) finds in Drosophila that the mutant
abdomen rotatum, in the fourth chromosome, has always a
dextral effect, whereas the similar third chromosome mutant
rotated abdomen is always sinistral according to Bridges and
Morgan (1923). In rats, King (1931) found an inherited
unilateral microphthalmia with a tendency for development
on the right side.

2. Another type of asymmetry is one in which an effect may
take place right or left or on both sides. This means that right
and left vary independently (no correlation) and that therefore
chance may produce the effect on neither side, on both sides, right
only, or left only. Such cases are characterized by a right-left
correlation coefficient nearing 0. For a number of Drosophila
mutations, such a condition has been described, e.g., by Guthrie
(1925) for eyeless, by Plunkett (1926) for achaete, by Astauroff
(1930) for tetraptera. The last author has made a special study
of the case of tetraptera. He concludes that here a variability
is involved which is more or less independant of the genie con-
stitution and of the environment but is caused by a local vari-
ability of the conditions of development. He works this out
in a very abstract way, but the facts could as well be described
by the action of an embryological process of the type that was
derived for the vestigial case. Astauroff thinks that the varia-
tions found in regard to the presence or absence of mirror sym-
metry in identical twins, as studied by numerous authors and
explained sometimes by somatic mutation in the case of absence
of symmetry (Newman, 1916), are also to be explained by
independent right-left variation.

3. Within a genie effect produced on both sides of the body, a
certain more or less considerable variation occurs independently
on both sides of the body. The pattern is intrinsically symmetric
but in detail asymmetric. We mentioned this type for the differ-

226 PHYSIOLOGICAL GENETICS

enl members of the vestigial scries in Drosophila (except the

lowest and highest) and for the color of the intersexual males of
Lymantria. In these eases, the interpretation is obvious: the
asymmetry is a product of the variability in detail of local
systemic conditions which regulate the flow of a growth sub-
stance (or other such stuff) spreading from the basis of the wing
(see page 222). In other cases, the same or a similar process
might furnish the explanation, and the authors who have studied
such eases have expressed themselves in a more general way
which might easily be converted into a picture of the same type
that we developed above. Thus, Haecker (1925) spoke of
variations in the internal conditions of the system. Wright
(1918), studying the asymmetries in color patterns of guinea
pigs, spoke of local factors controlling irregularities in develop-
ment. The same author (1935) mentions local conditions as
causing the asymmetrical expression of Polydactyly in guinea
pigs.

A different category within this group of phenomena may be
assigned to the hereditary asymmetry studied by Breitenbecher
(1925) in Bruchus (bean weevil). Here a mutant was found that
is inherited as a sex-controlled recessive (limited to females but
carried by both sexes) in which one elytrum carries two black
spots; the other, two red ones. Dextral and sinistral individuals
occur in equal numbers, so that it must be assumed that a
bifurcated chance process is decisive. It is very difficult to
formulate a proper explanation for this type of gene-controlled
asymmetry, except by going back to the first stages of develop-
ment of wings. (A parallel case in Drosophila is being investi-
gated by the author.)

The oldest example of this type wras already known to Darwin
and has been studied genetically by Przibram (1908). In cats,
individuals occur with one blue and one yellow eye. This
asymmetry is hereditary, and among the offspring of individuals
with blue on the right side individuals with blue on the left may
also occur. The inheritance of extra toes in poultry and guinea
pigs is of the same type. Many cases of this type occur in man
where much work has been done, as the problem becomes impor-
tant when identical twins are involved. Danforth, Newman,
Dahlberg, and others have discussed these problems and pre-
sented facts. Dahlberg (1929) has tried to formulate an explana-

THE MIT AT ED GENE 227

tion closely following Goldschinidt's .suggestions on the basis of a
chance distribution of formative materials.

There should finally be mentioned some recent non-genetic
experiments by Darby (1934) on the old problem of asymmetry
of claws in Crustacea. Przibram (1901) had shown that, after
removal of the large snap claw, at the next molt the pinch claw
produced a snap claw, and a pinch claw was formed from the
stump of the old snap claw. Much work has been done since
on the same and similar problems, without leading to an under-
standing of the processes involved. Darby removed first one
claw and then the other after different time intervals. Thus, he
found the critical time (about 30 to 42 hr. after the first operation)
at which the fate of the regenerate will be decided. By proper
experimentation he could thus shift the process so that on both
sides pinch claws or on both sides snap claws or on both sides
intermediates were formed. He refers this change of asym-
metry into symmetry to the relative times of production and to
the quantity of two morphogenetic substances A and B. The
factors controlling them he calls generally environment and points
out that here environment supersedes the effect of the genes.
From this he concludes that the genes are responsible for the
production of the substances, and the environment for their
quantity. If we substitute for environment the combination of
integrating processes, all of which are gene controlled in normal
development and all of which may be shifted by external agencies,
we come again to the point of view regarding symmetry and
asymmetry that we have developed above.

Ludwig (1936) has discussed the same experiments and
attempts an explanation which is different from Darby's (and a
similar one by Dawes, 1934). His argument is as follows. The
claws have hereditarily an alternative norm of reaction (pinch or
snap). In regeneration, the decision falls in favor of the alter-
native for which at this moment the larger quantity of determin-
ing material is available (see Lymnaeus, page 224). This
material is being increased because its production is controlled
by the type of claw that is developing. A formal explanation
for the rest of the experiments may be derived this way. Also,
this point of view may be expressed in the terms developed above.
But as no genetical facts are known that might be linked with the
experimental results, this short review may suffice.

22S PHYSIOLOGICAL GENETICS

3. Special Patterns Superimposed upon the General Pat-
tern (Secondary Tertiary, Etc., Patterns;. Two examples will
show what is meanl by the foregoing heading. The general
pattern of a Drosophila wing is given by the arrangement of
veins and spaces between the veins; form and arrangement of
hairs; and the typical proportions, shapes, and curvatures of the
wing. In the series of mutants of the vestigial or beadex type,
or similar series, this primary pattern of the wing is not affected
or only as far as certain regulations and readjustments are
necessitated locally as a consequence of the abnormal develop-
ment. But an additional new pattern is created, the typical
shapes of the partly destroyed, scalloped wings in different
members of the series. For a second example: in the gypsy
moth or Promethea moth, female and male wings have essentially
the same primary pattern of bands and spots. But in the
male, this pattern is overlaid by a dark pigment covering more
or less the primary pattern. We consider these types first
because they are simpler and also better known genetically.

If we look at a series of mutants of the vestigial series in
Drosophila, we find a definite pattern in each, e.g., the pattern
type Notched or Strap (see Fig. 19). These patterns are char-
acterized by the following features.

1. A certain regularity is observed if a series of stages is
compared. The scalloping begins at the tip of the wing; it
destroys large parts of the wing before the alula and posterior
cell are affected ; it destroys the central part of the wing along its
entire length last, beginning at the tip.

2. There is a certain irregularity: the details of the scalloping
are rather variable, and often no two wings exactly alike. The
two wings of one individual vary independently in regard to the
form of scalloping (not in regard to its amount), and symmetry is
reached only in the lowest and the highest grades of scalloping
(see, however, page 223). But for a given mutant type the
average amount of destroyed area is constant within certain
limits. We have already given the interpretation of the phe-
nomenon of scalloping as derived from the study of development :
an insufficiency of a substance needed for growth or the produc-
tion of a lytic substance above a certain threshold produces the
degeneration that leads to scalloping. If we consider, now, the
pattern according to which the different effects occur from nick

THE MUTATED GENE 229

up to vestigial, e.g., by drawing a typical series of these patterns
together into the contour of a normal wing (Fig. 41), we see a
way of understanding these patterns. If a definite amount of
the supposed growth substance diffuses into the wing from its
base — to use only one of the alternative explanations, the other
being its reciprocal — the path or bed of such a flow will be
determined by the structural, physical conditions of the organ,
and a more or less definite bed will be prescribed. If this flow
ceases at different times before the whole wing is covered (insuffi-

Fig. 41. — Wings of a series of v^-alleles drawn together to demonstrate the path
of the destructive process.

cient amount of the substance), the rest of the wing will degener-
ate, and the contour of the degenerating area, of the scallopings,
will represent the front line of the flow and the moment of its
being stopped. The whole series of stages of scalloping, then,
furnishes a cinematic picture of the bed along which the substance
in question spreads. The pattern produced is therefore a func-
tion of the amount of spreading growth substance controlled by a
mutant gene (or by a phenocopic effect) and of the structure
of the organ providing the drains and obstacles, shortly to be
described as the conditions of the system. In case of explanation
by spreading of a lytic substance, the moving picture would be
the reciprocal of the foregoing: the filling in of the wing area with
this substance, beginning at the tip.

This interpretation recalls at once a case that we have already
described, the pattern upon the wings of intersexual males of the
gypsy moth (Fig. 35). We remember that here, too, the facts
led to the conclusion that the determination of scales in the male

230 PHYSIOLOGICAL GENETICS

is produced l>y a stream oj a substance entering the wing from
its base and spreading in a radial direction without fixed beds but
not completely without order. Also, in this case, there was a
relation between the quantitative features of this flow and the
resulting pattern, lather the time allotted to the spreading
of the substance determines how much of the wing .surface
will be covered, £.< ., the relative amount of male and female area
or the amount of substance available is responsible for this. Ii
both cases, then, we have a rather irregular pattern effect of
different arrangement in detail, controlled by the action of genes
The genes in question do not, however, control the pattern
directly but reactions of definite speed, leading to production of
definite substances at definite times or in definite quantities
The diffusion of these substances into the organ — generally
described as a determination stream — in beds prescribed by the
si ructure of the organ, together with the time of onset or quantity
of the flow or both, produces in one way or another the resulting
pattern. There can be no doubt that this simplest method is the
most frequent one in general processes of formation of embryonic
patterns. An indication of this will be found in the fact, also
emphasized by Timofeeff (1934), that most of the mutants of
morphological characters in Drosophila tend to showr the type
of asymmetrical behavior just described for the vestigial case.
Color mutants, however, the genes for wrhich do not act through
induction of pattern formation, are symmetrical in effect. To
this should be added, as a nice check, the fact that vermilion eyes,
which we know (see page 186) to be formed partly by an inductive
process of the horraonic type, are frequently found in asymmetric
expression.

The same material may lead one step further. We have stated
that a series of facts similar to those concerning the vestigial
wing of Drosophila was found for the truncated wing (see Gold-
schmidt, 1937). In this case, the degeneration of tissue as a
result of insufficiency of some substance takes place below the
wing epidermis, which remains intact. The wring margin is then
retracted beyond this degenerating area (Fig. 42), caves in, and
forms the truncate wing, which also is known in a series of
increasing conditions, caused genetically as well as by phenocopic
effect. Here, now, we are facing a perfectly regular pattern,
viz., the dumpy shape with curved margin of the Drosophila wing

77/ E MUTATED GENE

231

caused by a gene-controlled process which is not directly con-
cerned with pattern but leads automatically to a pattern as a
consequence of the conditions found in the system, here the wins
Anlage.

A closer study of the wing pattern of the intersexual gypsy
moth leads to another insight. In spite of the general aspect of
lack of order in the flow of the determination stream, we find

Fig. 42. — Three stages in the development of a dumpy wing in Drosophila from
a normal pupal wing. {From Goldschmidt, 1937, Univ. of Calif. Publ. Zool.)

occasionally definite relations to other elements of the wing. It
is not surprising that frequently the wing veins may form the
edges of a pigment tongue. But more important is the fact that
sometimes a pigment streak ends along one of the zigzag bands
(Fig. 43) (see also Kuehn and Henke, page 198). The impression
is that these bands stopped the further progress of the stream.
In this case, it is only an occasional happening (see Minami,
1925). Conditions within the wing Anlage might easily be such
that a determination stream would always be stopped by a band,
i.e., by some part of the structure of the organ which is determined

232

PHYSIOLOGICAL GENETICS

Fig. 43. — -Outlines of wings of intersexual males in Lymantria dispar to demon-
strate the relation between color pattern and veins and bands. Female parte
black. {From Minami, 1925, Arch. Entwicklgmech. 104, Fig. 3.)

THE MUTATED GENE

233

independently. In this case, the same type of flow of a sub-
stance across the wing would lead to a perfectly regular pattern.
There can be no doubt that many patterns found in our models,
the wings of Lepidoptera, are of this type. Examples will be
discovered wherever mutant genes do not change the original
pattern of the species but superimpose upon it a process that
leads to a secondary pattern. This will frequently be the case
with sexual differences and also with mutants of the type of

Fig. 44. — Pupal wing of Papilio podalirius. The part of the wing to the
left of the undulated line will degenerate, leaving only the tail. {From Sueffert,
1929, Zeitschr. Morph. Oek. 14, Fig. 9.)

melanisms where a pigment covers the preexisting pattern in a
definite degree, which might be quantitatively graded by the
action of different allelomorphs or genes. A number of such
cases have been analyzed genetically (Goldschmidt 19216,
1924a; Federley, 1920; Kuehn and Henke, 1929-1936), but little
else is known that would permit going beyond these general
statements. The most important additional fact is that similar
patterns to those produced in these cases by genes may also be
produced by temperature action, e.g., melanic patterns of one
sex in Gastropacha (Standfuss, 1896), Lymantria (Kosminsky,
1909; Federley, 1907; Goldschmidt, 19226); general melanism in
Habrobracon (Schlottke, 1926); and Ephestia (Kuehn and
Henke, 1929-1936). In this group belong also such secondary

23 »

PHYSIOLOGICAL CENETICS

patterns as scries of ripple-like lines between the main elements
of the pattern. They may he partly the expression of such
embryonic structures as the network of blood sinuses, as suspected
by Goldschmidt (1920(2). But thus far this type of secondary

pattern has not yet been studied
genetically or experimentally. For a
genera] discussion, see Henke (1935).
One more case might be mentioned
in this connection which belongs
generally to the type represented by
the wing form of Drosophila. It is
known that many species of Papilio
have tails on their hind wings; it is
further known that in some species
only males have tails; and, further,
that in some of the polymorphic
species, females may have no tails,
short tails, or tails according to the
presence of different Mendelizing
genes (de Meijere, Fryer, etc.; see
Goldschmidt, 1931c/). A similar effect
may also be produced by action of
temperature (Standfuss) or shock

Fig. 45.— Relation between (Poultoil).1 Sueffert (1929a) showed
form and pigmentation of scales ' ,

in the flour moth. Eight types that the development of this wing
of pigmentation. IV and VI pattern — formation of tails — is such

form the dark bands in wild ,

type; VIII forms the light that first a wing of normal contour is
pattern elements. The others formed, from which the tails are cut

are found in the background.

(From Kuehn, 1936, Naturwiss. out through a secondary process of
24, Fig. 5.) degeneration (Fig. 44). It is needless

to discuss in detail the generalization that here a pattern-forming
process may be understood on the same lines as in our former
examples.

In describing the case of the pattern of intersexual wings, we
repeatedly mentioned the fact that male areas have scales of the
typical form for males and, similarly, female areas, female scales
in type and arrangement (Fig. 46). Here we are facing another
important point. One and the same process — the spreading

1 Poulton gave a different explanation of Van Someren's material; see
Goldschmidt, 193 Id, p. 475.

THE Mi TATE I) GENE

235

determination stream — leads to formation of pigment, eventually
to a formation of pattern and simultaneously to quite typical
details of morphogenesis, here in the development of scales.
Rather similar facts have later been found in the flour moth,
where a definite relation exists between pigmentation and form
of scales (see Kuehn, 1932; and Fig. 45). If we add that Feder-

1 i ~

Fig. 46. — The scales upon an intersexual mosaic spot in Lymantria dispar.
{From Poppelbaum, 1914, Zeitschr. ind. Abstl. 11.)

ley (1907) had produced typical changes of scale form by tem-
perature action, we see that we are facing the same series of
interrelated facts: gene action or phenocopic action— ^production
and definite course of determination stream— ^general and special
morphogenetic pattern. Thus, the series of facts mentioned in
this chapter and their logical connection furnishes an important
step toward the understanding of the action of genes in con-
trolling pattern.

At this juncture we might report upon the much discussed but
not yet transparent case of the scute allelomorphs in Drosophila.
This large series of alleles produces a definite pattern of bristles,
some of which have been discussed. Definite bristles are absent
in each case, but with a certain amount of variation. We shall
not mention here the implications of this case as far as the theory
of the gene is involved; that will be done in a later chapter. We

236

/'// YSIOLOGH 'AL OENETICS

are reporting only such parts of the discussion as have to do with
the production of the specific bristle pattern by the action of
genes, whatever they are. The school of Serebrovsky (Dubinin,
Lewit, Agol, etc.) which has furnished most of the data has used
them only in an attempt to analyze the gene itself (see page 295).

FlQ. 47. — Location of bristles of Drosophila mclanogaster. aor, mor, par,
anterior, middle, posterior orbitals; oc, ocellar; pv, post vertical; iv, ov, inner,
outer verticals; uh, Ih, upper and lower humerals; anp, pnp, anterior and posterior
notopleurals; ps, presutural; asa, psa, anterior, posterior supra-alar; apa, ppa,
anterior, posterior postalar; adc, pdc, anterior and posterior dorsocentrals; asc,
psc, anterior, posterior scutellars. (From Plunk ett, 1926, Jour. Exp. Zool. 46.)

Sturtevant and Sehultz (1931) and Goldschmidt (19316) tried to
look at the case from the standpoint of pattern formation, and
Child (1935, 1936) has recently attacked the problem also.

The work in question is based upon the idea that these different
alleles produce a definite pattern of bristles by removing specific
individual bristles. If this is done in an orderly way, the
phenotypes of the different alleles may be arranged in a corre-
sponding seriation. Figure 47 represents the bristles in question
on head, thorax, and scutcllum of Drosophila, with their respec-

THE MUTATED GENE

237

tive names. Figure 48 shows for 15 alleles which bristles are
affected by the respective genes. The fact that such an arrange-
ment is possible demonstrates a certain order in the scute effect,
i.e., a pattern, and in addition an orderly behavior of the different
patterns. This latter order is brought out in the experiments
of the Russian authors who analyzed all the possible compounds
of these alleles and found that the compound always showed only
an effect upon those bristles common to both types entering the

ho

mrdc

W sa

[j

ov utpnp

ps

anp

por pa aor mor oc

ex pv sc

st

h

12

W

1

2

3
4

• i

o 9

w 10
11
13
14
15

Fig. 48. — The step allelomorphism of the scute alleles. Abbrev. as in Fig. 47.
mr, microchaetes on thorax; vs, ventral bristles; ex, coxals; w, crumpled wings.
{From Dubinin and Friesen, 1932, Zeitschr. Biol. Zentralbl. 52.)

compound. This led to the arrangement as figured, which has
been called a step arrangement.

There is, of course, no sense in analyzing this case from the
standpoint of pattern production, if these primary facts do not
hold. They have been accepted as valid by Sturtevant-Schultz
(1931), though their seriation is a little different. In recent
papers, G. Child (1936), however, doubts if there is a problem of
pattern involved at all. He bases this conclusion upon the
following facts. In temperature experiments on one scute line,
each bristle reacts independently of every other; it is determined
independently for each bristle whether it will be present or
absent. Flies raised at 14° show variations of some bristles, the
same bristles being constant at higher temperatures, and vice
versa. Such flies raised at different temperatures appear as
different from each other as do two different scute alleles raised
at the same temperature. From this he concludes that the
appearance of a pattern is an illusion, produced by looking
at the statistical differences in mean numbers instead of at the
condition of the individual fly, in which the distribution of bristles
may differ from that of any other individual.

238 PHYSIOLOGICAL GENETICS

I do not think thai these arguments arc valid. We know from
all the cases presented above that by temperature effects the
phenotype of one type may be changed into that of a mutant
type (phenocopy). This does not mean, however, that no
pattern is involved in the production of those types but rather
that tin1 process of pattern formation is of such a nature that it
may be shitted in different directions through external (tempera-
ture) or internal ( modifiers) changes. The cases that have been
described illustrate this elementary principle. It is not an inte-
gral part of the pattern conception that a pattern-forming process
is invariable; it is, indeed just as variable as any other process,
and its laws will therefore become better visible in statistical
treatment than in the study of individual cases. Child himself
finds that at a temperature used by the Russian authors the
effect upon bristles agrees with the seriation given by these
authors. Apparently, a bristle pattern is actually involved,
though it may be shifted by modifying agencies.

A second point to be clarified is the type of pattern involved.
If we think of our former discussions regarding the outlet of the
determination stream (called center of diffusion in Plunkett,
1926, and Child's papers), the following different situations might
be found: (1) There is one point of outlet, and the whole pattern
in all cases (barring now the variations) is the consequence
of the conditions of the whole system prescribing the bed of the
determination stream. (2) There are a number of outlets, and
their combined workings with the general conditions of the
system in time (time of opening of the outlets, time of flow) and
space (direction of the stream, eddies, obstacles, etc.) will control
the resulting pattern. (3) Each bristle corresponds to a separate
point of outlet, comparable to the different typical spots in a
piebald pattern. Then there would be no space element, a flow,
but only a time element — opening of the outlet — and a threshold
element — sufficient quantity of determining stuff for bristle
formation (or bristle destruction). The first two possibilities
may be combined with the peculiarities met in our former cases:
the bed of the stream might be thoroughly fixed (many moths) ;
it might be variable within certain limits (the vestigial series);
and it might almost be left to chance (the intersexual wing of
Lymantria). In the last case, each individual would be different,
though a statistical treatment would reveal an average pattern.

THE MUTATED GENE 239

In the case considered as 3, there was no determination stream
that could have different beds. But the causative agent for the
functioning of each center of diffusion (at each bristle) might
change in regard to the time of the onset of action (or the thresh-
old), and there might be more or less order in this respect; this
would give the same phenotypic result as the different types of a
determination stream. Sturtevant and Schultz (1931) thought
of the type with one point of outlet. Goldschmidt (19316) tried
to derive from the facts a system with five such points. Dubinin
and Friesen (1932) as well as Child (1936) think, however, that
the actual variability of the behavior of the individual bristles
does not show the type of correlation that would be required if a
determination stream should coordinate the behavior of groups of
bristles that it would have to reach in a definite order. This
is a potent argument in favor of the possibility mentioned under 3.
The recent work of Nujdin (1936), at least in part, supports these
conclusions. He made a special study of the correlation of
different areas by comparing their behavior in mosaics, exhibiting
different areas containing bristle- and hair-affecting genes
(scute 8). It is to be remembered that the mesonotum (dorsal
thorax and scutellum) is derived from one pair of imaginal disks.
The sternopleura might develop from either the dorsal or the
ventral mesonotal disk, and the humeri are formed from an
independent disk. He found that all areas derived from different
disks behave independently. Within the mesonotum he finds
nine areas on each side that behave as units in regard to mosaicism.
This means that in the differentiation of the mesonotum at some
time a subdivision takes place which results in a pattern of the
surface of the mesonotum. According to Nuj din's description,
some fields in this pattern contain only one, and others two,
bristles (nine areas and 1 1 bristles) . This, then, means that most
of the bristles, but not all, are found in areas that have been
formed as independent parts of a pattern. It has to be added
that within a given area a stronger correlation exists in an
anterior-posterior direction. This means that the bristle pattern
is the result of a formation of areas in a definite pattern, a
formation that is bilaterally independent, proceeds generally in
an anterior-posterior direction, and takes place almost simul-
taneously. If this description is correct, one might conclude
that it is not a series of determination streams that are involved

240 PHYSIOLOGICAL GENETICS

bul a more or less simultaneous stratification of the whole field
into areas followed by the bristle-forming reaction. In a general
way, these results agree with those of Rokizky (1930).

The question is what this would mean for the problem of
bristle pattern. This pattern means the formation or not-
i'ormation of one (or two) bristles within certain areas. (It
might also be a destruction of bristles after their formation, as
suspected but not yet proved.) Child has found that all indi-
vidual bristles in a given mutant line have the same temperature-
effective period. This indicates that the bristle pattern is not
concerned with the moment of bristle formation, as expected from
the foregoing considerations. But there is an important item in
Child's last paper: in different mutants of the series there is a
difference in regard to the sensitive periods and to the time of
development.

This leads to the last part of the problem. Goldschmidt had
assumed that a determination stream is involved, but he was
careful to point out simultaneously that this assumption is not a
necessary part of the argument, which holds also in case of inde-
pendent determination of each bristle. The decisive point was to
explain the steplike arrangement of the phenotypic effects in the
series. This would be completely explained if the different
areas of the pattern in the mutants had a different time of
differentiation, the beginning as well as the duration of which is
controlled by the different alleles. Expressed in terms of
determining substances, this would mean that the outlets would
open at different times in the different areas or that the velocity
of differentiation became different in the areas or a similar process
involving a time element. Such a system, as worked out by
Goldschmidt in a model that might be changed in many directions
(see also Friesen and Dubinin who constructed a similar model in
order to criticize it), accounts sufficiently for all the facts in a
general way. The observation of Child, which was just men-
tioned, is highly in favor of such an explanation by relative rates
of some processes. We refrain, however, from going into more
details. If Child's (rather improbable) claim that the step order
of the Russian authors and also of Sturtevant-Schultz is of no
value were justified, any discussion would be superfluous. Future
studies will be necessary to clarify the problem.

THE MUTATED GENE

241

4. Composite Patterns. — Considering rather simple types of
patterns, we discussed cases where the limiting factor for the
spreading of a determining stuff was given by a pattern already
present in the system, e.g., the zigzag bands of the Lymantria
wing. The analysis of the effect of mutant genes will more
frequently meet with the simpler type of the processes because
Mendelizing genes affect more often late details of morphogenesis.
But analysis of more complicated cases of pattern has also
yielded considerable insight. Thus far, however, it has not

Fig. 49. — Pupal wings of intersexual males of Lymantria dispar after drying,
showing the finished white scales and the soft collapsed scales in the later dark
parts of the wing. {From Goldschmidt, 1923, Arch. mikr. Anat. 78.)

been possible to fit all the partial processes that have been found
into one simple coherent picture. The attack upon this problem
has been done entirely with the Lepidopteran wing as material,
except for a few facts relating to mammals. There are two types
of facts of primary importance which thus far could not be
properly linked, though this is required for a full understanding
in the study of the lepidopteran wing.

a. General Embryological Features. — In studying the develop-
ment of patterns in the lepidopteran wing, Goldschmidt (1920c?)
found one general rule which may be stated as follows: A long
time before pigment is formed and before the scales are perfected,
a different rate of differentiation becomes visible in regard to the

242

PHYSIOLOGICAL GENETICS

later pattern: parts that later will carry dark pigmented scales
show a slower differentiation of the scales than parts that will
later carry white or yellow or red scales (distinguished by con-

Fig. 50. — Pupal wings of the swallowtail, Thais polyxena. a, fully developed;
b, before pigment formation, later white parts with finished scales, later pig-
mented parts collapsed. (from Goldschmidt, Fig. 48.)

taining uric derivatives or carotinoids or air). This can be
detected only during a very short period of differentiation: if a
pupal wing at this stage is dried, the already differentiated scales
remain erect; the undifferentiated scales on later dark areas,
however, collapse because they are soft bags filled with blood.

THE MUTATED GENE 243

Thus, the main features of the future pattern become visible
in the form of a relief of protruding and of flattened scales.
Figure 49 shows such a stage from an intersexual wing, the later
male parts being represented by the valleys of the relief. Figure
50 shows a similar condition in the wing of a swallowtail butterfly,
compared with the finished pattern; and Fig. 51, a similar stage
in the formation of the eyespot in a Saturnid moth. This
differential velocity of differentiation within different parts of the
pattern was found in numerous different objects from diversified
groups,1 including also the zigzag bands of Lymantria. It was,

Fiu. 51. — Dried pupal wing of Samia cecropia. Region of the later eyespot
with crescent of white scales, which are finished and erect. (From Goldschmidt,
1920, Arch. Entwicklgmech. 47.)

however, not observed in Kuehn's work on the flour moth, the
reason probably being that the very quick development of this
form makes it difficult to find the short stage when the phenome-
non may be made visible.2

In the flour moth, moreover, Kuehn's students Koehler (1932-
1935) and Braun (1936) found another nearly related fact. At
an early stage of development of the wing, the place of the future
zigzag bands is indicated by a zone of more numerous mitoses.
Two such waves of increased mitotic activity spread across the
wing: one in the hypodermal tissue; a second, in the scale-
forming cells. If a mutant (or a phenocopy) has a different
arrangement of the bands (shifting toward the base or the tip),
the pattern of mitoses, indicating the later dark areas, is corre-
spondingly shifted (Fig. 52). In addition, it was found that the

1 Mostly unpublished work by Goldschmidt and Sueffert.

2 It has been found meanwhile by W. Braun in the author's laboratory.

244

PHYSIOLOGICAL GENETICS

number of scales formed in the later dark areas is larger per unit
than in the light areas. A difference in the relative increase of the
number of scales in light and dark areas was not found. These
facts are in perfect harmony with Goldschmidt's discoveries (a
point not realized by the flour-moth investigators) ; t he occurrence
of many mitoses means that the cells in question have not yet

(a)

(b)

CO

Fig. 52. — Pattern of mitoses in young pupal wings of the flour moth, 66ft. a.
normal; b, shift toward the center in the mutant with narrow field of symmetry.
(From Braun, 1936, Arch. Entwicklgmcch. 135, Fig. 7.)

reached the end of their multiplication period, after which
growth and differentiation (formation of scales) begins. They
are therefore retarded in differentiation as compared with the
cells that do not divide further. Another fact has been found by
Koehler which falls in line: in the flour moth the wing rudiment
folds up at a certain stage, and these folds coincide roughly with
the situation of the future zigzag bands. This, again, agrees with
Goldschmidt's observation, as the softer, less differentiated
parts of the pattern, located in the zigzag bands, which will still
undergo mitotic division, are the natural lines of folding.

We spoke of different speed of differentiation of the scales in
the different areas of the pattern. This must not be mistaken
for different growth. This is roughly identical all over the wing.

THE MUTATED GENE 245

The difference is actually found in the velocity of hardening of
the chitin. As pigment is deposited in the chitin in colloidal
solution, a hardened scale cannot be pigmented any more, and
therefore the speed of hardening, different in different areas,
leads to the pattern of pigmentation. Recent work in the
author's laboratory has confirmed these basic facts, which there-
fore cannot be overlooked any longer.

These two groups of facts together show that a primary pattern
has been formed early in development for the whole wing as a
unit. The pattern consists in the creation of zones of differential
velocity of differentiation, which later will lead in some way to
different pigmentation and sometimes shape and structure of the
scales. This unit pattern is influenced by mutant genes. (Sexual
difference might be included in this statement, because genetically
the difference is of the general type of a mutational difference.)
Can we now conceive of the gene-controlled process that leads
to the formation of such a complex pattern? In the case of the
intersexual wing, we have found the answer: here the pattern of
areas of different speed of differentiation was conditioned by the
determination stream spreading from the wing base across the
wing. There is no reason to suppose that the process is a
different one (in principle) in the other cases. What has to be
explained, obviously, is why the production and spreading of
such a substance leads to induction of definite growth velocities
only at definite points or, using an expression previously intro-
duced, why this induction becomes stratified.

b. Pattern Localization. — In Goldschmidt's first publication
on the subject (19206), a rather generalized formulation was
applied to this problem. The special arrangement of the parts
of the pattern was referred to as the arrangement of the outlets of
pigment, which now has to be defined as the arrangement of
points from which the determination stream starts. It was
known from the genetic analysis of different patterns that a
number of genes may be concerned in controlling this arrange-
ment. Forthwith an attempt was made (Goldschmidt 1920d) to
change experimentally the processes involved by injuring the
young wing in different ways, setting defects, etc. But only the
very beginnings of success were attained. As is now known from
Henke's work (1933a) upon the same material and using partly
the same methods, the lack of complete success was due to

246

PHYSIOLOGICAL GENETICS

ignorance regarding the actual situation of the sensitive period,

after the end of which the pattern cannot be changed. In recent
years, Kuehn, Henke, and their .students have successfully
attacked this problem.

These experiments are partly based upon an interesting prog-
ress in the study of the morphology of such patterns, due to
Schwanwitsch (1924-1929) and Sueffert (1925-1929). These

3-

A
/*\

1 1
/ I

2-
1-

/
/

/ \
/ \
/ \

/ \

\

a x

<t

/ W"

-1-

Vo / V / -

2-

0.05 0.10 0.15
H r1 r1 1 ' i ' i ' i

2-

1-

n

/ v

-1-

6 \ "Q y^ _ o^Vj

■2-

V

0.05 0.10

r^ r-1- 1 h 1 H p

0 12 24 36

60 72 84 % 0 12 24 36

60 72

(a) (b)

Fig. 53. — Sensitive periods for the dark pattern elements in flour-moth pupae.
R, number of dark scales in the marginal spots; M, in the central spot; Q, the
dark bands; S, the "shadows." Abscissa: age of pupa in days and in fractions
of whole length of pupal period. Ordinate: units of the controls. O = M of
the controls. {From Kuehn and Henke, 1936, Abh. Ges. wiss. Goettingen, 15,
Fig. 87.)

authors showed that within a given group of Lepidoptera (nym-
phalids, Schwanwitsch, Sueffert ; saturnids, Henke (1928, 19336, c,
1936), the wing pattern may be reduced to a simple arrangement
of elements that are invariably found. Figure 36 showed diagram-
matically such a case; the central field of symmetry with the
two symmetrical bands is noticeable. Outside lies the ocellar
system, followed by the marginal bands. Near the roots of the
wing are the so-called hollow elements.

Sueffert (1925) and Kuehn (1927) further showed that in
temperature experiments these different "fields" reacted inde-

THE MUTATED GENE 247

pendently: a definite treatment may change only the field of
symmetry or another field. In addition, the temperature-
effective period may be situated at different points of develop-
ment (Koehler and Feldotto, 1935, Wulkopf, 1936). Figure 53
shows the curves for the degree of expression of different elements
of the pattern in the flour moth after heat treatment at different
times of development; the maximum effect marks the sensitive
period. In addition, Kuehn and Henke found in the flour moth
a number of mutants which affected definite parts of this system.
We have mentioned one that enlarges the symmetry field. Such
mutants had also been known for other objects. Thus, Gold-
schmidt had found a recessive mutant in the gypsy moth, in which
all the zigzag bands belonging to the so-called symmetry system
had disappeared; another that affected only the marginal
markings; another affecting the symmetry field but not the zigzag
bands; and others affecting general color independent of pattern.
A case of a similar type is the color pattern of coccinellids. In
Epilachnia chrysomelina, Tenenbaum (1933) showed that many
details of the individual spots are controlled independently by
different genes. The next step was made when Henke 1933a
succeeded in changing details of pattern by operating upon the
young wing.

By cutting off parts of the imaginal disks or burning parts
before final determination had taken place, Henke succeeded in
shifting elements of the pattern in the Cynthia moth. The
bands of the symmetry system are here somewhat differently
arranged on fore- and hind wing. Proper operations may
change each of these types into the other one. The formation
of the bands after operation showed clearly that they are formed
at the borders of the central field, i.e., at the edge of a determina-
tion stream, as illustrated before for the flour moth. Also, other
elements of the pattern could be shifted; e.g., the crescent-like
spots could be subdivided into two; the marginal bands could be
changed. Though every detail may not yet find a proper
explanation, it is shown, in general, that the process of pattern
formation occurs by movement of substances in definite paths
which are set by the conditions of the whole system and may be
changed by removal or destruction of parts.

Similar experiments by Kuehn and associates upon the wing
of the flour moth and of Abraxas have already been reported, and

_' is pb YSIOLOQICAL GENETICS

we saw how they demonstrated the existence of a determination
stream, comparable to the one discovered in the Lymantria wing.
All these facts now lead to the decisive question: How can the
action of the genes be linked with these different aspects of
pattern localization? In his detailed analysis of the problem,
Goldschmidl (1927c) stated thus the general facts and viewpoints
that have to be taken into account whenever this problem is
being attacked:

1. The pattern of a lepidopteran wing is formed by the definite
arrangement of areas, which contain scales different in color and
partly in structure. These areas may be chemically different,
e.g., contain melanin, guanin, carotin, etc. They may be differ-
ent only in regard to quantity, e.g., contain different concentra-
tions or grades of oxidation (or reduction) of melanin. They
may be different physically in regard to the form of scales or their
surface structure (optical colors).

2. Viewed as a whole, the patterns result from combination of
rather few elements: specifically colored or structured scales;
zigzag bands; broad or narrow bands which are transversal,
oblique, or parallel to the veins; points, specks, spots in definite
relations to veins or wing margin; systems of wavy lines, con-
centric lines (eyespots); areas without scales.

3. Within single families, e.g., Sphingids, Papilionids, Arctiids,
the majority of species show patterns that consist of numerous
permutations of the same general plan of arrangement. (These
plans have been worked out for different groups by Sueffert,
Scfrwanwitsch, and Henke).

4. Patterns of very similar type may be different in regard to
chemical or physical structure; the same element may be char-
acterized in one species by chemical, in another by optical colors,
etc.

These rather general facts already offer some points of attack
for an analysis. As the pattern is generally composed of rela-
tively few elements, differing in different species mostly in a
quantitative way, it is to be supposed that the processes of
pattern formation are also of a relatively simple type, which,
however, permits of many permutations and quantitative shifts.
Furthermore, since all patterns show definite spatial relations
to the areas and general form of the organ (wing base, edge, etc.),
the formation of pattern must be connected in some way with the

THE MUTATED GENE 249

conditions within the wing as a unit. Other typical relations
of the pattern to definite parts of the wing, such as the veins, are
known, and these special structures also must be a part of what
have been called the systemic conditions for pattern formation.
Since a general plan of pattern is found in the different larger
groups, different systemic conditions are supposed to be present
in these groups. Finally, since similar or even identical patterns
may be combined with different chemical or physical structures
of an area — or, vice versa, the same chemical behavior over
different areas — pattern formation and chemical specificity, e.g.,
guanin, melanin, carotin, as pigments, must be caused by inde-
pendent processes of determination.

To these general points of view two more special ones of general
importance have been added.

5. Forms are known in which different generations (alternate
spring-autumn or wet-dry seasons) exhibit different patterns,
which sometimes, in tropical forms, may be so different that they
can hardly be conceived of as simple quantitative shifts within a
pattern; they actually show a different plan of pattern. Exter-
nal conditions acting upon the same genotype may then produce
extraordinary changes of the type of pattern. (They can be pro-
duced also experimentally; see the classic case of Vanessa levana-
prorsa). In view of what precedes, it may be inferred that the
embryonic causation of such very different patterns may be due
to relatively simple differences in the causative agents.

6. Cases are known of extreme polymorphism within a species,
e.g., the many different forms of females of Papilio dardanus,
mimicking the patterns of very different families. In such cases,
it is known that a few Mendelian genes are responsible for the
entire array of patterns. It follows that extraordinary differ-
ences of pattern may be produced by the simple differences of
reactions due to a mutant gene. In the same case, the different
patterns are further controlled by sex; they are limited to one
sex. The differences in development of the two sexes therefore
may permit or prevent the realization of different types of pat-
tern. This shows, again, that a relatively simple alternative
may lead to far-reaching consequences regarding the pattern.

To this general review of the situation we may add a list— in
no way complete — of known effects of mutant genes upon pattern
in Lepidoptera:

250 PHY8IOLOOICAL GENETICS

1. The chemism of all areas of one type, < .g., between the dark
dements of pattern, is affect eel. The mutant gene acts at a
level above the pattern, which is not changed. All mutants,
mostly recessive, sometimes sex linked, changing a red-ground
color to yellow or a yellow color to white are of this category.
Examples: red and yellow Callimorpha dominula (Standfuss,
1890; Goldschmidt, 1924a); yellow and white Colias (Gerould,
1923); Parasemia (Standfuss, 1897).

2. Within the given pattern, all elements containing melanin
pigment increase simultaneously and restrict the nonmelanic
areas. These cases of melanism usually are of the polymorphic
type, i.e., typical quantitative steps from beginning to complete
melanism are found and a series of additive (polymeric) genes
are involved in the production of the different steps. Examples :
Melanic polymorphism in C. dominula (Goldschmidt, 1924a);
Spilosoma lubricipeda, (Federley, 1920; Goldschmidt, 1924a).

3. Within a given pattern, the whole pattern is intensified or
bleached. Examples: Abraxas grossulariata lacticolor (Don-
caster); Argynnis paphia-valesina (Goldschmidt and Fischer,
1922). In these cases, one sex-linked (Abraxas), viz., one sex-
controlled, gene (Argynnis) is involved.

4. Within the given pattern, individual elements, e.g., the
so-called middle shadow within the symmetry field, increase in
regard to melanic scales.

Recessive or dominant genes may be involved and also sex-
controlled behavior. If different genes controlling the same
process but for different elements of the pattern are involved,
combined and additive effects are produced by combination
of these genes but in a different way from the foregoing group.
Examples: Lymantria monacha (Goldschmidt, 19216); flour moth
(Kuehn and Henke, 1929-1936).

5. The whole unchanged pattern is overlaid by a melanic
pigment. Here belong the frequent simple cases of recessive
melanism, also of sex-controlled melanism. Example: many
geometrids (T. W. H. Harrison, 1920). Here belong also the
cases in which the pattern is unchanged but the whole tint of
coloration is changed in degrees from light to dark melanins.
Example : the male wings of geographic races of Lymantria dispar
(Machida, 1924; Goldschmidt, 19336). Here a series of multiple
alleles is involved.

THE MUTATED GENE 251

6. The basic pattern itself is affected by a shift of the relative
distance (or size) of the elements. Example: enlargement or
restriction of the field of symmetry in the flour moth (Kuehn and
Henke, 1929-1936). A dominant mutation is involved here.

7. The basic pattern is affected by suppression of some of its
elements. Example: the recessive mutant lunata of L. dispar,
(Goldschmidt, 1927c, 19336) showing no zigzag bands in the
system of symmetry.

8. The whole basic pattern is changed into a quite different
pattern. (Examples: mimetic Papilios, De Meijere, Fryer,
Poulton, and collaborators; see Ford, 1936), where the process is
sex controlled and dependent upon a few recessive genes and their
recombination.

These are the types of genetic facts, which are to be correlated
with the embryological facts in order to gain a general insight
into the all-important problem of the production of pattern by the
action of genes. It may be added that a large bulk of similar
genetic evidence could be cited regarding the control of pattern
by mutant genes in the caterpillars of Bombyx and Lymantria
(Toyama, 1906a; Tanaka 1913, 1916; Goldschmidt, 19216,
19286) ; in the larvae and imagos of chrysomelid beetles (Tower,
1906; Tenenbaum et al., 1933). But in all these cases, the addi-
tional experimental and embryological knowledge are missing.

There is little difficulty in explaining all these cases in which
genes control differences within the pattern (change from red to
yellow, melanism overlaying pattern). The action of the
mutant gene is of the same type as has been described. It does
not involve the problem of pattern itself. Goldschmidt (1927c)
worked out a theory designed to reduce this problem to the
already known principles regarding action of the genes. His
interpretation may be stated in general terms thus: The genes
controlling pattern act by producing definite, reactions of definite
velocity. These are properly timed; i.e., their time elements
are arranged so that the thresholds of effect are reached in a
definite order, thus insuring for each process a definite situation
upon which to act. In the known cases where there is genie
control of the velocity of differentiation in different areas of the
pattern, at a given moment two such areas are found in a different
stage of development. If, now, another gene-controlled reaction
leads to the production of, say, guanin, to be deposited in the

252 PHYSIOLOGICAL GENETICS

scales, only those scales can be impregnated with it which are
in the proper stage. This means that the substance in question
i- deposited only in definite areas. (See the facts of scale
development, page 241.) This one model serves for the descrip-
tion of many other types of patterning in similar terms, using
velocities, thresholds, time of onset or end of action, and similar
processes of simple quantitative behavior as elements. This has
been worked out in Goldschmidt's book (1927c).

But the decisive point remains: This or any other explanation
requires that a primary stratification occurs at a definite time
leading to areas of different constitution. Only then can begin
the automatic action of a system, as the one described.1 These
primary areas in our case were the areas of different speeds of
differentiation (caused primarily, so it seems, by different rates
of cell division, see page 244). Goldschmidt assumed that this
process occurs as a consequence of the production of a substance
at a definite time, which acts upon the substratum, i.e., the young
wing, like the evocator substance in the formation of Liesegang
rings within a colloidal medium. This means that the genes
responsible for this primary pattern or stratification act just as all
genes do, producing a definite substance at a definite time; but
that "the conditions of the system," i.e., the physicochemical
structure of the Anlage and the laws of chemical equilibrium,
cause a reaction with this evocator substance which leads to an
automatic distribution of the products of reaction in a definite
arrangement or pattern, after the model of the formation of
Liesegang rings (see discussion, page 205). On the assumption
that the whole wing acts as a unit, there was no possibility
of making more specific assumptions regarding this phenomenon
of stratification as applied to the pattern of the butterfly wing.
The work of Henke and Kuehn (1929-1936) has led one step
further. They showed that in the formation of the primary
pattern, the following processes take part.

1. An already present general subdivision of the wing area into
different fields, which are, however, not yet determined in their
later extent. In the terms that we used, this would mean that
there is not one single process of evocation for the whole wing
but a number of centers from which this process starts. (These

1 This is, of course, the same situation that was discussed repeatedly in
regard to embryonic pattern.

THE MUTATED GENE 253

are the outlets as designated in Goldschmidt's earlier work.) It
would also mean that the Liesegang-ring model would probably
not apply to this part of the problem.

2. For at least one of these fields it was proved (as reported
above) that its limits, the later system of symmetrical bands, are
formed at the edges of a determination stream which flows with
known speed across the wing in a definite bed (see Fig. 38) and
at a definite time. Here, then, we actually see the process of
stratification, caused by a spread of evocator substances from
different points of outlet and causing the deposition of some
product of reaction (here a substance retarding differentiation or
producing mitosis) where the stream comes to a standstill (or
meets the neighboring field). The decisive points of Gold-
schmidt's interpretation have thus not only been found to be
correct, but they have, in addition, been put on a safe experi-
mental basis. This important step taken by Henke and Kuehn
permits the following application of general ideas about genie
action to further detail of pattern formation (to be found partly
in Goldschmidt's first discussion, 19206). A number of quantita-
tive variables are added to the system, which can account for
many details of pattern formation by simple quantitative shifts
in one or another direction. These are the relative time of
opening of the "outlets" of the determination stream into
different fields (see the different sensitive periods); the relative
velocity of spread of the determination stream as determined
by its own composition and the physical nature of the sub-
stratum; the relative amount of substance available as a deter-
mination stream; the amount of reaction product and the locus
where it is formed or accumulated. Needless to say, mutant
genes are those which act by quantitatively changing one of these
variables.

Some of these have actually been analyzed by Kuehn and
Henke in the mutants and phenocopies of the flour moth. The
results of temperature action in the sensitive period in different
lines were compared with the results of operation upon the wing.
Some of the details that have not been mentioned thus far are the
following. In a pupa of 72 to 84 hr. of age, the determination of
the pattern has been completed, resulting in the described process
of increase of number of mitoses in the later dark areas. Two
periods are distinguishable within the first 3 days of pupal life

254 PHYSIOLOGICAL GENETICS

(at 18°C). In the first, a temperature shock enlarges the sym-
metry field; the operation shows that the determination is not
yet finished and that the stream is in progress. In the second
half of the 3-day period, temperature shock produces narrowing
of the field in question. The operations (destruction of parts of
the field) produce the same effect and show in 48 to 60 hr. old
pupae the determination stream in flow, as reported. It starts
from both wing margins, joins in the center and then spreads
toward wing base and tip. A stopping of this flow means a
smaller field. One of the mutant genes Sy stops this process of
spreading before it is finished. Certain steps of this spreading
process are more frequently realized in the different experiments,
which might indicate areas of special resistance to the spreading
(or a wavelike process of spreading, Goldschmidt); one such
line coincides with the limits of the fields produced by the gene Sy.
There is another gene Syb, which broadens the field of symmetry;
this might also be done by temperature shock. The two effects
cannot be added (see the different behavior of the vestigial case,
page 22), and therefore the authors assume a line of absolute
resistance in the substratum. Here, then, have been analyzed
two phases that might influence the pattern by quantitative
variations and their mutual interplay: (1) what the authors call
the impetus of spreading, which is the same as the quantity of
determining substance in Goldschmidt's old interpretation; (2)
the resistance to spreading, which in Goldschmidt's terminology
was the conditions of the system. Since it is found that opera-
tions upon AS?/-animals more frequently fit one definite stage of
the spreading of the determination stream, as compared with the
same operations upon Wild type, it is specifically concluded that
the gene Sy does not affect the stream but the resistance of the
substratum, preferably at a definite time. These and similar
facts then demonstrate clearly the presence of a system of such a
type as had been postulated by Goldschmidt in his first attacks
upon the problem.

• As all these details are rather complicated, we might finally give
a short picture of the origin of any more complicated type of wing
pattern, adding that each of the steps may be controlled by
mutant genes, which shift the process in question quantitatively
in time of onset, duration of action, locus of action, quantity
of reaction product, and threshold of action. The first pattern

THE MUTATED GENE 255

to be laid down very early in development and not yet analyzed
genetically or embryologically is the pattern of wing shape and
distribution of wing veins. This primary pattern is of very
early origin (Goldschmidt, 1920d, drew attention to the fact
that the pattern of wing veins is already visible on the chitin
of the pupa before these veins have been formed); and it fixes
the systemic conditions for the later processes, resistances,
drains, etc., as previously described. The second pattern is the
pattern of future fields or areas determined by the location of
the points of outlet of a determination stream or field centers.
Extreme cases, such as the case of Papilio dardanus, show that
this pattern is controlled and may be shifted by mutant genes
(also sex differences) considerably. The genetic facts point out
that very simple events, like shifts of a process in regard to time,
may be involved in such changes, but nothing is known yet as to
how these points are determined primarily. There is a suspicion
that they may coincide with growth centers. The third pattern-
ing process, described below, controls the limits of these areas or
fields and the type of pattern of visible field limits (bands, etc.).
This process is the formation of a determination stream proceed-
ing from these outlets with definite velocity and in a bed set by
the foregoing patterning processes (systemic conditions). Time
of onset and length of flow of this stream, together with corre-
sponding states in the neighboring areas, and the systemic
conditions, determine the detail. Mutant genes are known to
change these processes, also temperature effects. The fourth
step found is the production of differential cell divisions at the
edge of the stream, which thus appears to be of the type of a
growth substance. Differential mitotic patterns again lead to
an earlier start of the phase of scale differentiation outside the
edges of the stream, i.e., later cell differentiation at the edges, the
bands. The latter cell groups are those which will receive pig-
ment by being in the proper condition for its colloidal solution,
when another process supplies the chromogen. This pattern,
again, may be overlaid by a new gene-controlled process which fur-
nishes more or less melanin, or different-colored substances, to
all or any fields that are in a condition to receive them. Again,
mutant genes control these productions.

c. Additional Facts. — In order clearly to present those facts
which have pushed the analysis to the point where it stands

256 PHYSIOLOGICAL GENETICS

today, we have repeatedly passed over points tluit would have
required more consideration. We propose to discuss these now in
the form of additions.

The decisive point, the analysis of which is still more or less
in the theoretical stage, is the formation of those primary patterns
which have been described as the distribution of the field centers
or points of outlet. This problem is, of course, identical with
the problem of embryonic stratification, as mentioned before.
We have often used the comparison with the production of
Liesegang rings in colloidal solutions after introduction of some
chemical, acting as an evocator. Gebhardt (1912) was the first
to draw attention to the great similarity of some patterns on
butterfly wings with Liesegang rings, and he believed that pig-
ment deposition in the scales was actually such a process.
Kuester (1913) applied this same principle to numerous rhythmi-
cal structures in plants. In its original form, the idea had to be
discarded as applied to the butterfly wing, when it became known
that the pattern was determined independently of pigment
formation (Goldschmidt, 1920d), though it was still possible to
make use of the same idea in explaining the arrangement of
different determining stuffs (growth substances) in the form of a
rhythmic pattern or stratification. As we have seen, this is
probably not the case for the primary patterns of the butterfly
wing, though certain secondary patterns within the fields or
spreading over the whole wing may be explained thus. (Simi-
larly with many rhythmic developmental patterns, like segmenta-
tion.) A few additional facts regarding this phenomenon may
therefore be mentioned. In detail, it may be produced by differ-
ent processes. A substance present in the system may spread
by diffusion and be arranged in a rhythmic way; or the phases
of the rhythm may be produced only as a reaction in loco; or a
precipitate may be formed in a rhythmic procession. We may
assume that one or the other of these phenomena occurs when
rhythmical patterns are involved, and we may accept this as a
general model for developmental, simultaneous stratification
in cases where no other related process of diffusion is known.
Thus far, no actual genetic or experimental fact is known to
show a Liesegang phenomenon at work, though some pat-
terns apparently require it. Quite recently, however, Becker
(1937) showed that a special pattern formed upon the tergites

THE MUTATED GENE 257

of wasp queens, independent of the typical pattern of wasps,
is the product of a Liesegang ring formation with mechanical
evocation.

There are rhythmical patterns which suggest at first sight a
Liesegang phenomenon. Henke (see 1935) calls them simul-
taneous rhythms and points out that they are characterized by
the absence of a center of diffusion and therefore must have a
different origin. Here belong the tiger and zebra stripes and their
like. Henke points to the fact that tiger and lion hybrids exhibit
a pattern that is a kind of compromise between the stripes of a
tiger and the rosettes of a young lion, which shows that the
different genes of these species control the same general process.
Many authors have tried to find an explanation for such patterns.
Haecker was the first to assume that rhythmic growth of the
skin was responsible, in so far as pigment would be deposited in
zones of intensive growth. This, of course, would refer the prob-
lem to the formation of a primary growth pattern. Goldschmidt
(1920c) for some time accepted this solution but was later led
to a different formulation when he had seen that a young tiger
embryo exhibited the whole later pattern in form of epithelial
papillae, which represented the Anlagen of the large hair (Leit-
haare). This same fact was found and elaborated by Toldt (1912)
in a series of papers. This pointed to a primary pattern forma-
tion in relation to the arrangement of hair follicles. Krieg (1922),
who returned to this subject, thinks of tensions within the skin
at the time of pattern determination ( = distribution of some-
thing) which makes the determining stuffs collect in definite
lines. This makes it possible, of course, by adding the rate
concept, to explain different patterns by different times of onset
of the action of determining stuffs. This view, indeed, accounts
well for the facts and has the advantage of permitting tests in
forms like dogs, cats, and cattle, where these patterns are con-
trolled by mutant genes, and other patterns by other mutant
genes. Such an analysis is still missing. Growth tensions, of
course, would then adequately explain a primary, general, and
simultaneous pattern, not showing any centers of origin. As such
tensions would be the automatic result of former developmental
processes, another type of automatic patterning could be added
to those already discussed. Henke (1935) points out that many
elementary processes of pattern formation might come into this

_>;>s PHYSIOLOGICAL GENETICS

category, e.g., the arrangement of hair, feathers, glands, etc.,
in the skin.

In studying the wing pattern of Lepidoptera, we had to point
out the existence of primarily determined centers (outlets) for
the determination process, the centers thus dividing the surface
of the organ into fields. The same problem occurs again in the
piebald patterns of mammals (also fishes). Goldschmidt
(1920/0 pointed out that there the same two phenomena are
involved as in the moth wing, viz., the inherited pattern of outlets
(field centers) and the control of quantity, berth, time of flow,
etc., of the spreading substance responsible for pigmentation.
Kuelm (1926) and Henke (1933a) have shown that the fields
thus created are rather independent of each other; e.g., A may
be black, and B white, or vice versa; in A, very little pigment
may have "flown out" from the centers; in B, none; and in C,
t he whole field may be covered. This points to a process with a
time element, which Goldschmidt (1920c) had termed the order
of opening of the outlets. Genetic work with such patterns
shows that mutant genes exist, controlling the quantity of released
pigment, frequently in an additive way in series of multiple
allelomorphs; further, genes controlling the arrangement of the
centers and their symmetry; also, genes controlling what we called
the order of opening of the outlets, resulting in special features
for each field; and, of course, also the genes controlling the color
of pigments (see the numerous papers on spotting in mice,
rabbits, rats, guinea pigs, horses, dogs, by Castle, Wright, Dunn,
Phillips. Ibsen, and others). The material already existing in
these groups might profitably be used for a deeper analysis of
this type of pattern, an analysis that has not yet been made from
the standpoint of physiological genetics. Attention may be
called to Iljin's study (1926) of piebald patterns in guinea pigs.
He assumes that centers of pigmentation (the outlets) are not
found, but centers of depigmentation, i.e., the negative picture.
This leaves, however, the problem of gene-controlled pattern
where it was. The facts reported above (page 155) regarding
Himalayan rabbits and Siamese cats, demonstrating an interest-
ing interaction between gene products and environment, ought
also to be kept in mind, because internal conditions might assume
the role of environment and superimpose a pattern effect upon
a nonpatterned gene effect.

THE MUTATED GENE 259

A beginning of such an analysis has been made for a similar
object, the feather areas (pterylae) of birds, which certainly
again present areas with a center of outlet of feather-determining
stuff. Within such an area, according to Holmes (1935), first
a row of follicles are formed simultaneously; from here the process
spreads laterally and medially as well as anteriorly and posteriorly
and might therefore be described as a determination stream.
Within such a field, certain regularities have been found by Juhn
and Fraps (1934); Juhn, Faulkner, and Gustavson (1931); and
Fraps and Juhn (1936). The feathers on both sides of the first
row may behave as mirror images in regard to structure and color;
their rates of growth and length of regenerated shaft have a
definite size increasing in an anterior-posterior direction. Fraps
and Juhn showed a similar relation for barb length in regeneration
relative to the distance of the feather from the central row and
were able to describe this in terms of field functions (which is
only a description in mathematical terminology for the same
things which we treated thus far as spread of stuffs). These
pterylae coincide partly with areas of pigmentation (Juhn),
and possibly the same processes are involved in their formation.
Since here, again, mutant genes are known controlling the
different features of the pattern, we may expect more information
from this material.

Another attack upon this problem has been made by Goodrich
(1933), in fishes. We have referred to his work on Oryzias where
he could show that black and yellow melanophores are involved.
In yellow fish, the melanophores are present but not able to
form pigment. As the Dopa reaction shows, the oxidase is not
lacking, but the chromogen which is produced by the action
of the dominant gene. The difference between the two alleles
is thus one of quantity of effect. This effect takes place in the
individual cells; it is cell localized. As these chromatophores
migrate in the embryo from a center of origin, their determination
in regard to chromogen formation occurs probably in this early
stage. There is also a variegated type which requires that in
the early stage a pattern has already been formed of cells with
different powers of chromogen formation. But this pattern could
just as well be determined later by local conditions. No experi-
ments of the type of Twitty's (see page 179) have yet been
performed to answer this question.

260 PHYSIOLOGICAL GENETICS

A real pattern process is involved in the goldfish (Carassins
auratus). Here black and yellow chromatophores arc present,

and the mottled patterns arc caused by a secondary destruction
of pigment cells in definite areas (Berndt, 1925; Fukui, 1927,
1930). The resulting type depends upon the time of beginning
of destruction, place and amount of destruction, and whether or
not also yellow chromatophores are finally destroyed. Thus a
seriation brown self — variegated black and orange — orange —
variegated orange and white — white may be obtained. The
pattern itself is quite irregular and is complicated by the presence
of a pattern of scales with and without (transparent-type)
guanin crystals. There is, according to Fukui, a relation between
the pattern of chromatophore destruction and the structure
of the skin, and the destruction of pigment is combined with
phagocytic action. Goodrich and Hansen (1931), who made a
quantitative study of these facts, point out that most of them
may be explained by a system of reactions, producing the destruc-
tive agent with definite velocities. In this respect, we may
compare the case to the case of destruction of the vestigial wing.
But here, as there, remains the problem of pattern. In the
vg-c&se, it was obviously a consequence of the structure of the
wing, prescribing the bed of spread of the lytic substance. In
the goldfish, the pattern seems to be imposed by the distribution
of lymph spaces from which the lytic action starts, a theory
actually assumed by Fukui. Goodrich points out that the
pattern in Oryzias and in Carassius is to be explained on a
different basis. There is no doubt that the latter is of the type
resulting from the interaction of a gene-controlled process
(production of the lytic substance) with what we termed the
conditions of the system. (We might apply the terminology
used before and speak also of the arrangement of the outlets for
the lytic stuff.) The principles obviously are always the same,
though the details may be very different. In the case of Oryzias,
we do not know the patterning agent, but no doubt it will fall
into one of the known categories. It may be added finally that
experiments in transplantation of scales (Goodrich and Nichols,
1933) show that the lytic action is limited to definite times in
development, as expected.

In other fish species, more complicated patterns appear. In
Platypoecilus, for example, Gordon and Fraser (1931) have traced

THE MUTATED GENE 261

a number of pattern elements to individual genes. As these
patterns may be combined in a single individual, just as such
wing patterns in Drosophila as miniature, beadex, expanded
may be combined in one wing, each element must be caused
by an independent embryological process. Therefore, in
these and similar cases (Lebistes, Winge), an analysis will
have to be applied of the type used for the wing of the flour
moth. No facts are yet known to warrant more specific
conclusions.

We may finally give the list of types of pattern formation
as compiled recently by Henke (1936), adding such information
as is pertinent to our topic :

1. Patterns based on chimeric development. These are the
patterns in gynandromorphs, chimeras, somatic mutations.
They throw light on problems of autonomous differentiation but
hardly on actual developmental patterns.

2. Patterns based on rhythms in time. Here belong the
patterns of feathers, produced by hormone injection. Possibly
also patterns like feather pattern in Plymouth Rock. Their
relation to genie action is not yet clear.

3. Crack patterns. These have not yet been mentioned.
Regular cracks, as in a shellac film or craquele china, have been
used by Hirata (1935) to explain regular patterns in beans. If
we substitute zones of different growth or tension for the cracks,
the consequences will be the same as for the latter pattern-form-
ing processes.

4. Areas of diffusion from one center. Here belong the cases
of spotting, pterylae and their like.

5. Field limits as pattern. Here belong the best analyzed
cases of the bands on the lepidopteran wing, determined by the
front of a determination stream. They are combined with a
pattern of differential velocity of differentiation.

6. Rhythmic diffusion from centers. This is the Liesegang
ring type of pattern-forming process.

7. Simultaneous rhythms. An instantaneous rhythmical ar-
rangement, e.g., by tensions in a membrane, producing zebra-
like patterns.

8. Combination patterns. This is the Lepidopteran type of
pattern, composed of different areas with a combination of some
of the foregoing pattern-forming processes.

262 PHYSIOLOGICAL GENETICS

9. Pattern by differential growth. This is not contained in
Henke's review, because nol involved in color pattern, but most
important for embryological pattern formation. All of the fore-
going types may combine in the embryonic patterning, though
some of them may be more frequent, viz., 4, 5, 6, 9.

There can be no doubt that the formation of patterns, i.e.,
the decisive process of development, may, even in the present
stage of our knowledge, be reduced to the action of a gene-
controlled system of properly timed reaction velocities. These
reactions set in motion the different types of stratification
processes discussed and simultaneously provide the details by
proper distribution of onset, length of action, and threshold of all
the other incidental processes. A simple event at a definite
time and place will automatically set in action the whole series
of consecutive events. But each partial event may be shifted
by mutant genes or the action of modifiers or the environment.
There is no reason to believe that other developmental patterns
are controlled in a different wray from those which have been
used as models.

III. THE CYTOPLASM AND THE ACTIVATION OF THE

GENE

In our discussion of the facts and interpretations of physiologi-
cal genetics, we were concerned mainly with the action of the
genes, situated within the chromosomes of the nucleus. We
took for granted that the seat of the reactions controlled by
the genes is the cytoplasm, both the cytoplasm of the individual
cells and the cytoplasm of a group of cells taken together as a
unity (a field or territory in the sense of experimental embry-
ology). The role of cytoplasmic differentiation in development
and determination has been discussed in connection with the
phenomena of stratification, primary chemodifferentiation, and
different types of induction. In all these cases, the cytoplasm
was considered to be the substratum upon which the genes
work, though the processes happening within the cytoplasm may
take their independent course once started by the action of the
genes. In this chapter, we have to consider the facts that may
lead to a better understanding of the part assigned to the cyto-
plasm in heredity.

The questions that will have to be answered are:

1. If the cytoplasm is the substratum upon which the genes are
acting, is it possible to formulate more specific ideas of this
interaction?

2. At the outset of such an interaction, which must be of an
orderly sequence, a process is supposed to occur that may be
called the activation of the gene. What does this mean?

3. Is the activity of the cytoplasm in conditioning orderly
development always steered by the genes, or is an independent
activity of the cytoplasm known?

4. If the cytoplasm has independent genetic properties, are
these of the same type as the properties controlled by genie action,
or are they of a different, possibly more generalized, type?

1. INTERACTION OF NUCLEUS AND CYTOPLASM

The first question, regarding the type of interaction of genes
and cytoplasm, is difficult to answer, because only very indirect

263

264 PHYSIOLOGICAL GENETICS

evidence is available. In :t general way, a number of possibilities
are given. There is, first, the following alternative: Either the

genie action, whatever it is, takes place within the resting nucleus,
or it is confined to the mitotic stage, when chromosomes and
cytoplasm are in direct contact. In favor of the latter possibility,
the facts may he mentioned regarding the relation of mitotic
divisions to determinative processes (see page 244). But these
facts may also be interpreted differently. In favor of the first
alternative, the behavior of egg cells of the determinative
type may be mentioned. In such cells, decisive developmental
processes, the chemodiffercntiation, occur in the resting nuclei.1
Still a third possibility might connect the two alternatives, viz.,
that the products of genie action are formed within the nucleus
but set free only when the nuclear membrane disappears. It is
known that at least in some eggs determinative processes are
started with the bursting of the germinal vesicle.

Goldschmidt (1927c) and Koltzoff (1935a) have suggested a
connection between this phenomenon and the liberation of gene-
controlled substances formed during the growth period of the
nucleus. In a way, such views go back even to the discussions
upon the meaning of the nuclear and chromosomal behavior
in the auxocyte, centering around the terms idioehromatin and
trophochromatin (see Rueckert, 1899; Goldschmidt, 1904). In
a general way, it may then be assumed that the interaction of
genes and cytoplasm occurs through the nuclear membrane,
though other possibilities cannot be denied. This might mean
either of two things: (1) The direct products of genie action
diffuse through the nuclear membrane. As these products are
probably in the nature of catalyzers, either only low-molecular
catalyzers are involved, or the nuclear membrane must be
permeable for large molecules. (2) Substances from the cyto-
plasm enter the nucleus, and only the products of catalysis
return into the cytoplasm. There are only very few facts
available on which to base inferences. Most suggestive are the
strange and complicated intranuclear processes which occur in
the growing ovocytes of Selaehia and Amphibia and which
demonstrate an immense activity of the chromosomes, connected

1 Quite recently Stern (Nature, Oct. 30, 1937) has drawn attention to the
fact that genie actions are known which occur within single cells during the
resting stage of the nucleus.

CYTOPLASM AND ACTIVATION OF THE GENE 265

with the elaboration of manifold substances, having to do
obviously with the growth and probably also with the chemo-
differentiation of the cytoplasm. A direct demonstration of
the relations is available only in the experiments of Haemmerling
(page 192), which show that the intact nucleus controls the
production of formative substances, showing a gradient away
from the nucleus. But we do not know how and where they or
their precursors are elaborated under the control of the nucleus.
Thus we must be content to know that certain decisive cyto-
plasmic activities and differentiations are controlled by the
nucleus, i.e., by the genes, without being able yet to follow
chemically the details of these interrelations. (A theory to
account for the facts of development in rather the opposite way
from the one usually assumed by geneticists has been published
recently by Just, 1936).

It may be added that Koltzoff (1935) thinks that some sub-
stances that are difficult to synthesize are actually identical
with genes, wrhich produce their like by assimilation and give
off the surplus into the cytoplasm. Davenport (1935) thinks
that the genes lie near the surface of the nucleus. They attract
cytoplasmic molecules with the opposite charge. The residual
charge suffices to attract another molecule to the second, resulting
in making the two cytoplasmic molecules independent of the gene-
molecule, which starts the same process over again. It must be
honestly said that thus far this problem has not found any
factual or theoretical solution.

2. THE ACTIVATION OF THE GENES

The second question concerns the activation of the genes.
If genes control definite cytoplasmic processes occurring at
definite times in development, e.g., the production and flow of
the organizer substance in amphibian development, they are
supposed to be activated at the proper moment to set in motion
the cytoplasmic process. We have already reported the many
facts that demonstrate the time of onset of genie action in
different cases, ranging from the very end of differentiation
down to the behavior of the first segmentation spindle. In
order to form a general idea of these processes of activation, we
may describe them as follows. Whether the genes are catalyzers
or not, the decisive immediate product of their action must be

260

PHYSIOLOGICAL 0ENETIC8

catalyzers. As enzymes are extremely specific, their action
requires t h«- proper substratum; to which might be added the
optimum conditions in acidity, in regard to coferments, in the
relations of enzyme and carrier substance, in colloidal and
viscosity conditions of the medium. When all these conditions
are properly provided, the action of the enzyme (which might
always have I >< «ii present since fertilization, though perhaps not
in sufficient concentration) may begin, and this we call now the

Gene A
Gene B
Gene C

CADI ChDM Ch.D.m

B A C

Fig. 54. — Diagrammatic representation of the activation of the gene by proper
cytoplasm. {From Goldschmidt, 1927, Phys. Ther. d. Vererbg.)

activation of the gene. It was shown in the chapter on develop-
mental mechanics that development, from the standpoint of the
whole embryo, i.e., the cytoplasm, consists in a series of "stratifi-
cations" of cytoplasmic differences initiated by a gene-controlled
process. Thus, one substratum is transformed into two or more
different ones, which now can provide the proper substratum for
the action of new genes. We may thus describe the activation
of the genes in terms of the accompanying diagram (Fig. 54).
We assume three genes A, B, C which, or the products of which,
are ready for action as soon as the proper substratum is provided.
At a stage of development, Ch. DI, a cytoplasmic segregation
into two different substrata occurs. The lower one is the proper
basis for the action of gene B, and the chain of reactions controlled
by this gene is set in motion. At the stage Ch DII, the stratifica-
tion has produced five areas, the topmost of which is the proper
substratum for the gene A. The third stratification provides
the substratum for C. The actual facts may be much more
complicated, but it is most probable that the general relations

CYTOPLASM AND ACTIVATION OF THE GENE 267

between genie action and cytoplasmic component are of this
type. It has to be added that the situation might be obscured
if genie action had produced the pattern of stratification already
before fertilization. In this case, only the maternal genes would
have taken part in the process, which thus may appear as an
independent cytoplasmic action. This situation was discussed
on page 206.

3. CYTOPLASMIC HEREDITY

The third problem is the one usually meant when the role of the
cytoplasm in heredity is discussed. Is the protoplasm not only
the substratum for the action of the genes but also an independent
part of the germ plasm, directly influencing the hereditary traits
by action of its own?

A. Merogony

It is known that some of the classic work in experimental
embryology was directed to a solution of this problem (Boveri,
Godlevski). In the experiments on merogony (purely paternal
nucleus in the cytoplasm of another species), first a positive
result was obtained, but this was later discarded by Boveri
himself as due to misinterpretation. In recent years, these
experiments have been revived with partly different methods.
Using the same method as Boveri, Hoerstadius (1936) produced
hybrid merogons between two species of sea urchins which could
be raised to the pluteus stage. The skeleton turned out to be
not typical. The club-shaped ends of the main calcareous rods
characteristic for one of the species (Psammechinus) are deter-
mined by the nucleus. They are intermediate in the hybrid and
typical in merogons with Ps. nucleus. In merogons with Ps.
plasma, abnormal rods appear; Hoerstadius, however, seems to
believe that this does not mean a plasmatic determination but
only an abnormality produced by the nuclear action in foreign
cytoplasm.

The second method applied to such studies is the production
of germinal layer chimeras (von Ubisch, 1936; Hoerstadius, 1936)
in echinoderms by transplanting the micromeres of one species
into the blastula of another. In these cases, combination
skeletons may be formed which look very much like a hybrid

268 PHYSIOLOGICAL GENETICS

skeleton. No influence of the cytoplasm could be found, though
it would be difficult to prove it.

The third method has been introduced by Baltzer (1933) on
the basis of old observations by Spemann, producing merogons
in Triturus. The Triturus merogons did not grow long enough,
however, to permit an analysis of the specific differences. Later,
it was shows that merogonic tissues, if transplanted to a normal
host, may survive and develop far beyond the stage possible in
the pure merogon. This fact has been used recently by Hadorn
(1936) to perform an experiment of the type discussed here. The
egg of T. palmatus is fertilized by sperm of T. cristatus, and then
the female nucleus is removed with a pipette. After gastrulation,
a piece of presumptive epidermis is removed and transplanted
to a gastrula of T. alpestris. The merogonic epidermis will cover
most of the left side of the resulting chimera. This can be bred
beyond metamorphosis. The skin of T. palmatus which furnished
the plasma of the merogon is characterized by papillae made up
of a series of flat cells. The epidermis of the part of the chimera
'derived from the merogon shows exactly this structure, i.e.,
the structure of the species that furnished only the cytoplasm
of the merogon. It must be added, however, that the alpestris
host produces the same structure. It is therefore possible that
the effect has not been produced by the cytoplasm of palmatus
but by an induction from the alpestris host, which, however,
would be a quite unusual feature (though existing in Aceta-
bularia). In any case, the experiment is not convincing enough
to derive conclusions upon cytoplasmic inheritance. Very little
is known of merogony in plants. East (1934) cites a few cases
obtained more or less fortuitously after crossing by Kostoff (1929),
Clausen and Goodspeed (1925), Clausen and Lammertz (1929),
and Nawaschin (1927), none showing any cytoplasmic influence.

A second somewhat similar type of experimentation in this field
has been performed on cells of fungi by Harder (1927). In
Basidiomycetes, a stage occurs after conjugation in which one
of the cells contains a mixture of the protoplasm of both parents
but only one nucleus. This cell may be isolated and grown.
If the two parents belonged to different forms (Schizophyllum
commune and Phuliota mutabilis), the influence of the mixed
protoplasm as contrasted to the one nucleus could be studied.
Some of the hereditary traits showed only the influence of the

CYTOPLASM AND ACTIVATION OF THE GENE 269

nucleus; but others, noticeably the growth habit, showed a cyto-
plasmic influence. This experiment ought to be repeated with
the necessary variations to make sure that actually a cytoplasmic
effect was involved. It seems probable that it was, in view of
the work soon to be reported.

B. Plastids

In all these cases, only the general effects of nucleus and
cytoplasm were approached by the type of experimentation,
without a possibility of isolating the action of individual genes in
different protoplasm.

Now genetic literature is overflowing with analyses of recipro-
cal crosses in which the nuclei are alike in regard to individual
genes, whereas the protoplasm is different according to the
mother. In the overwhelming majority of cases in which the
crosses are made between the wild form and its mutations or
mutations inter se, no influence of the plasmatic constitution
upon the result becomes visible. Reciprocal crosses and their
offspring behave alike, of course, with the exception of differences
in the X- or Y-chromosomes. We exclude also such differences
as find on closer inspection a special explanation. There are,
for example, the cases of plastid characters in plants, first studied
by Correns, which show inheritance through the mother, as
usually no plastids are transmitted by the pollen tube.

We do not intend to go into the details of the plastid problem
here, as we are concerned only with the problem of cytoplasmic
heredity. Among the cases studied by Correns (1928) are those
in which he came to the conclusion that plastids are prevented
from becoming green when situated in a definite cytoplasm, called
by him a diseased cytoplasm. But such cases might also find a
different explanation. Renner had shown (1924) that a normal
behavior of plastids may depend upon definite genetic constitu-
tion, as, in Oenothera crosses, plastids within their own cytoplasm
will not form chlorophyll in the presence of foreign chromosomes.
Following up this work, he came recently (1936) to the conclusion
that in all the manifold facts concerning plastid inheritance and
behavior, including differences in reciprocal crosses, no cyto-
plasmic effect is involved. His opinion is that plastids are
independent hereditary units which might even show changes
comparable to mutations. The facts of all the different types of

270 PHYSIOLOGICAL GENETICS

panschure arc attributed thus exclusively to the constant prop-
erties of the plastids themselves. If Renner's views prove to be
correct, the facts regarding plastid inheritance will be ruled out
of the problem of cytoplasmic action.

('. Cytoplasmic Influence upon Gene-controlled
Charactkrs

It is very significant that though no cytoplasmic influence upon
the action of genes in controlling visible characters has been
found, when simple mutants were involved in the cross — one
exception will be mentioned later — quite a number of cases with
positive results are available that are all alike in that the sys-
tematic distance between the parents is rather large. If we
omit at present the matroclinous species crosses which always
have been suspected of being caused by cytoplasmic differences
of the parents, the first properly analyzed case was found by
Goldschmidt (19246). The Mendelian characters in question
were differences between subspecies or geographic races of
Lymantria dispar, viz., amounts of pigmentation in the cater-
pillars of different races. The type of heredity found may be
indicated by the following diagram which applies to practically
all other cases.

Range of Variation

P, dark

1\ %ht,

F,. dark X light

Fi, light X dark

F2, (dark X light)2

F2, (light X dark)2

This means that Fi is matroclinous and that in F2 a 1:2:1
segregation takes place, the range of which, however, is shifted
in the reciprocal F2 toward the maternal side. A corresponding
effect was obtained also in backcrosses. The conclusion, then,
was that it is not irrelevant for the action of the genes on which
or within which cytoplasm they are acting. Within the two
races, the respective phenotype was the result of the action of
the genes, given the specific cytoplasm. If the same genes,
however, were working within a different cytoplasm, the result
of their action was shifted toward the type to which the cytoplasm
belongs. Light genes in dark cytoplasm — to use these brief

CYTOPLASM AND ACTIVATION OF THE GENE 271

designations — produce also in extracted homozygous condition
a phenotype that is darker than the pure light race, and, vice
versa, dark genes in light protoplasm fail to produce the com-
pletely dark forms among the homozygous segregants. When
.the same experiments were performed with different races dis-
tinguished by the presence of less and less light allelomorphs
of the same dark gene, the same result was obtained but in a
lesser degree. There can be no doubt that the geographic races
in question were different in regard to some element of their
plasmatic constitution. Since that time, other quantitative,
gene-controlled characters in the same material were found to
behave exactly in the same way, e.g., length of hibernation and
velocity of larval development (see review in Goldschmidt,
19346).

In recent years, a number of observations of a similar type of
inheritance have been made, and with attention drawn to this
problem more cases will be available. I mention only the report
by Murray and Little (19356) who found that hereditary tumor
incidence in mice, studied in reciprocal Fi and F2, behaves exactly
as the case just reported, thus showing an influence of the cyto-
plasmic condition.

In this piece of work, however, one important check was
missing, viz., the analysis of the extracted homozygotes in further
generations, which could not be made for technical reasons.
But similar work has been done since with similar results and the
additional filling of the just-mentioned gap, at least in part.
We shall soon see, however, that a decisive point is thus far
not settled. Of first importance in this connection is Wett-
stein's (1924-1934) work with mosses. In this material, crosses
between races, species, genera, and even subfamilies are possible,
whatever this may mean among bryologists. Generally speak-
ing, the crosses between races of one species were reciprocally
alike and did not show any influence of the cytoplasm upon the
gene-controlled characters. All other combinations, however,
showed such an influence, increasing with the taxonomic distance.
Among the characters studied, some, like the form of leaves,
showed very clearly the plasmatic influence after the manner
that we have just discussed for the Lymantria case; others showed
it as less pronounced or not at all. In the case of the leaf char-
acter, it was possible in the species cross to obtain a further back-

272 PHYSIOLOGICAL GENETICS

cioss generation which behaved according to the expectation
derived from the given explanation. In the generic cross, no
such direct way of testing was possible on account of the sterility.
But it was to he expected that among those spores which are
normally formed, because the respective gene recombinations
are viable, no difference of viability would exist between gene
recombinations of predominantly paternal or maternal origin.
As a matter of fact, more maternal recombinations were found
in the reciprocal crosses, and this is interpreted as meaning
that the genes work more successfully with the cytoplasm of
their own species. In this cross, it was also possible to backcross
over again in succeeding generations with the paternal species,
but the result was always maternal offspring, which shows that
the supposed plasmatic incompatibility with the foreign chromo-
somes has not been changed. Unfortunately, also, in this case
the decisive experiment with extracted homozygotes in recipro-
cally different cytoplasm could not be performed, except for
one generation of a backcross on one side only (see page 275).

The plasmatic influence that was found in these more com-
plicated crosses was tested in many different ways by Wettstein
and his students (Doerries-Rueger, 1929; Becker, 1931; Melchers,
1935) and was found to apply to many individual characters like
osmotic pressure and division rate. All these cases may, however,
be interpreted as a general action of the cytoplasm upon gene-
controlled reactions. No influence has been found that could
properly be called a cytoplasmic determination of a definite
trait.

One more case of a similar type in plants might be mentioned,
because it shows again certain of the general features that we
have met thus far. Sirks (1931a, b, c) found in his crosses with
beans no reciprocal difference of crosses within the two sections
of vicia faba called major and minor. But in crosses between
these sections, differences came to light that have to be regarded
as of plasmatic nature. If a certain linkage group of genes is
represented in a cross, the homozygous forms appear only in F2
if the genes in question are situated in the protoplasm of the form
from which they entered the cross. Otherwise, they are not
viable. In other words, the protoplasm of the major group is
lethal for a certain chromosome derived in homozygous form
from a minor plant. Or, stated another way, the chromosome

CYTOPLASM AND ACTIVATION OF THE GENE 273

in question contains cither a lethal gene or is as a whole lethal,
the gene acting only in major cytoplasm in a homozygous state.
A case studied by Honing in Canna subspecies crosses is similar
but less extreme in plasmatic effect; others found by Skalinska
and Gairdner fall in line. In addition to this case, still more
cases were found that resemble our Lymantria case in that not
only Fi in reciprocal crosses is different but also the segregants
in F2 are carried in one direction by the cytoplasm. In our case,
this direction was the direction toward the maternal type from
which the cytoplasm sprang: light cytoplasm, for example, made
all gene combinations lighter. In Sirks' case, however, the minor
cytoplasm generally increased the action of certain growth genes
but decreased that of others; similarly with color genes. This
difference between the two cases cannot, however, be considered
as of primary importance. The subspecies of Lymantria, which
were used, are distinct natural forms, isolated from each
other and therefore presumably well balanced in themselves
through long selection. The subspecies of the beans, however,
are cultural varieties and therefore probably not so distinctively
balanced. It must be stated that, unfortunately, in Sirks's
cases, the analysis of further extracted generations is missing.

We mentioned previously the fact that all the cases of plasmatic
differences thus far known and recognizable by genetic methods
were found in crosses involving at least subspecies. This does
not mean, however, that all such crosses must give this result.
Baur (1932), for example, reports that in his numerous crosses
between geographic so-called species of Antirrhinum, no recipro-
cal differences were found. But as this material has not yet
been worked up quantitatively, later work might change this
statement. There has come to my attention, however, only a
single case in which positive results have been found in an ordinary
cross with mutations. Kuehn (1927) reported the isolation of a
strain of the wasp Habrobracon differing in intensity of body
pigmentation from the stock. In crossing this form to the type,
he reports results that practically repeat our results in Lymantria
and are represented sufficiently by the diagram given above, with
the exception of multiple genes being involved.

We have dwelt thus far mostly upon cases in which the con-
clusions regarding plasmatic influence upon the gene-controlled
phenotype could be tested at least in the F2 generation. But

274 PHYSIOLOGICAL GENETICS

there is an additional set of facts derived from species crosses in
plants, where no Fj generation is available. Most of this work

has been done in crosses between Epilobium species by Kenner.
Lehmann, Schwemmle, and Michaelis (see Michaelis, 1933-1936).
We shall try to isolate the decisive facts from the considerable
body of details, facts that, however, are more nearly related to
the experiments in merogony than to those discussed thus tar
and still complicated by the fact that one of the characters that
has been studied is pollen sterility. (A more detailed review and
discussion are found in East, 1934.) Crosses between different
species may be reciprocally different. If they are different.
certain quantitative characters are always involved, which appear
to be inhibited in one cross. But within one species are also
found different biotypes which show this reciprocally different
inhibition after crossing. If a hybrid is backcrossed to the
paternal species, and this is repeated over and over again in
succeeding generations, finally plants are secured that contain
practically only paternal chromatin within maternal cytoplasm.
All their species characters resemble, in fact, those of the paternal
species, but the inhibitory action on quantitative characters, like
corolla length and pollen fertility, remains constantly on the
maternal side, as in Fi.

It should be added as a remarkable fact that the cytoplasmic
influence is stronger upon the haploid characters. Pollen fertility
reacts much more strongly to a foreign cytoplasm than sporo-
phyte characters (see Michaelis and Wertz, 1935). No doubt
a plasmatic influence upon gene-controlled characters is visible
in this case. Later, we shall mention some additional details.

Let us try now to find out the general features of the facts just
reported. The following conclusions seem to be warranted:

1. A considerable number of known cases showr that the action
of the genes in controlling hereditary traits is different, when the
same genes are working within a different cytoplasmic sub-
stratum. In some cases, this is known for definite genes; in
others, only for the whole genie outfit.

2. The action of the cytoplasm might consist in shifting the
genotype toward the type of the parent that has furnished the
cytoplasm, i.e., the mother. It might also consist in inhibiting
the action of single genes or of all genes, if they are situated
within a definite foreign cytoplasm; this inhibition might act

CYTOPLASM AND ACTIVATION OF THE GENE 275

only if the genes or whole chromosomes in question are present in
the homozygous condition, but in other cases also in the hetero-
zygous one.

3. Very few, if any, cases are known in which this differential
action of the cytoplasm is found in crossing different varieties or
mutants of one species. But the cytoplasmic difference appears
rather frequently if subspecies and still more distant forms are
crossed. As far as evidence goes, it seems that its effect upon
the action of the genes increases with taxonomic distance.

4. In one and the same case, not all gene-controlled characters
show the cytoplasmic influence. If we leave out of consideration
general actions of inhibition and also complete incompatibility,
it seems that such characters are most likely to be affected by
the cytoplasmic background, which are of a quantitative nature —
size of organs or parts, amount of pigment, shapes of parts, and
similar characters.

It would be important to know whether or not specific prop-
erties of the cytoplasm exist (aside from the serological specifity)
that are different in forms with a cytoplasmic action of the type
just considered. Bellingshausen (1935) attacked this problem
in Michaelis' Epilobium material. The protoplasm of E.
luteum turned out to be more permeable to potassium chloride,
glycerin, and succinimide than hirsutum plasm. The permea-
bility for urea is decreased by chloral hydrate in a different way
in the two species, a fact that is accounted for by a different size
of pores and the presence of different lipoids in the two species.
These properties remained constant in 12 generations of back-
crosses and are therefore actually cytoplasmic. The interesting
observation of Michaelis (1935) belongs here, that E. luteum
is resistant to mildew; hirsutum, not; while luteum plasma with
hirsutum genom are almost resistent. This difference might,
of course, be of a serological order.

D. The Theory of the Plasmon

This last point now leads to the question whether or not we can
derive any generalization regarding the cytoplasmic action under
discussion. Apparently, some botanists are inclined to assume
that certain hereditary qualities may be transmitted by the
cytoplasm in the same way as other characters are determined
by the genes within the chromosomes (Winkler, 1923; Renner,

276 PHYSIOLOGICAL GENETICS

L929). And the introduction of the term plavmon by Wettstein
for the entity of cytoplasmic net ion in heredity may be considered
an outward sign of such a tendency. I cannot refrain from
expressing my opinion that the facts do not encourage such a
view (see also I last, 1934). Where it was possible in the just-
quoted cases to base the analysis upon individual gene-controlled
traits, it was shown that no cytoplasmic heredity comes into play
but only a differential action of different plasmata as substratum
upon the action of the genes in controlling the differentiation of
definite characters. I do not think that a single case is known
of a definite hereditary character that is determined as such by
the hereditary constitution of the protoplasm, aside, of course,
from the special case of plastid characters. In only one way can
I conceive of the facts that have been reported: I assume that
genes act in controlling the differentiation of hereditary traits
through the control of chains of reaction of definite velocity.
Whatever these reactions may be in individual cases, one thing
may be regarded as certain; viz., their velocities will be dependent
among other things upon the substratum in which they take
place, i.e., the cytoplasm. The specificity of the cytoplasm in
question may be physical, e.g., a matter of viscosity or permea-
bility; or it may be chemical like the pH situation; but in any
case there is no reason to assume that it is of a type similar to
the difference between two genes. It is then not a plasmon,
acting in some way as an independent part of the germ plasm,
which is responsible for the phenomena in question, but simply a
physical or chemical property of the cytoplasm of the species.
Its most probable action is the inhibition or, in other cases, aug-
mentation of velocities of gene-controlled developmental reac-
tions, resulting in visible influences of cytoplasmic constitution
upon such characters, which are influenced predominantly in
their development by relative speeds of differentiation. In
addition, of course, a kind of poisoning effect of wrrong cytoplasmic
surroundings might be involved in cases of sterility and the like.

E. Cytoplasmic Inheritance and Dauermodification

Arrived at this point, we have to return to a fact that wre have
mentioned repeatedly, viz., the incomplete knowledge of the
behavior of the cytoplasmic influence in further generations of
extracted homozygotes. The importance of this point for a final

CYTOPLASM AND ACTIVATION OF THE GENE 277

solution of the problem will become clear if we turn now to the
consideration of a possibility of interpretation that has recently
been tried rather successfully.

We have mentioned before the fact that cases are known in
which the genes obviously influence the behavior of the cytoplasm
of the egg, e.g., the case of the dextral and sinistral snails. We
must consider, therefore, at this point the question whether or
not after all the differences of cytoplasmic nature under discus-
sion are not themselves the products of genie action. Of course,
such a simple relation is out of the question because this would
result in a type of inheritance like Toyama's so-called maternal
inheritance. But there is also another possibility. Jollos
(1921, 1924) has discovered the strange phenomenon of Dauer-
modification, in which the action of external conditions changes
certain characters of an organism and the change reappears in
the next generation in a lesser degree and may continue thus for a
number of generations until it finally disappears. A number of
such cases in Protozoa and Metazoa are known (see Haemmer-
ling's review, 1929), and in all of them the obvious explanation
is that the change in question is a cytoplasmic one, which only
slowly, in the course of a few generations, is overcome by the
action of the genes, controlling the normal cytoplasmic constitu-
tion. The experiments designed to test this interpretation have
given consistent results (Jollos), though it should be added that
all decisive tests have not yet been performed. Jollos and
Haemmerling have pointed to the near relation of this body of
facts to those reported earlier. If an experimentally induced
Dauermodification may need up to 8 or 10 generations to dis-
appear, one might conclude that in the previously mentioned
experiments of Michaelis and Wettstein no proof is contained that
the cytoplasmic difference between the two forms, which was
originally present, remains in spite of the presence of foreign
genes. It might be quite possible that longer continued experi-
ments would show that, just as in Dauermodification the effect
of the stimulus disappears in the course of some generations,
so the primary, natural plasmatic differences will slowly disappear
under the influence of the genes of another form. This would
mean that the primary plasmatic differences of subspecies and
species would also be controlled by the genie constitution but
would be of such a nature that one series of ontogenetic processes

278 PHYSIOLOGICAL CENETICS

under the control of different genes would not suffice to change it.
As a matter of fact, experiments arc available to test this problem.
The method is to backcross a hybrid that shows maternal
influence over and over again with the father, so that the genes
are finally mostly paternal, but the cytoplasm is still derived
from the original mother. Wettstein performed this with the
species crosses of mosses for a number of generations without
finding a change in the cytoplasmic influence. Michaelis, how-
ever, who performed the same experiment in Epilobium, found
that all the characters that showed the cytoplasmic influence
continued to do it after eight generations of paternal backcrossing
but with a clearly measurable reversion toward the paternal type.
This, of course, favors the views of Jollos, which we just
represented.

A similar effect, visible even in F2, has been found by Honing
(1930) in tobacco crosses, involving hereditary reactions to
light. In a more recent paper (Michaelis and Wirtz, 1935),
Michaelis minimizes his former results and claims that in spite
of a certain reversion the luteum cytoplasm has remained typical
after 13 generations with a hirsutum genom. There can be no
doubt that the decisive work in this field has still to be done.

F. Cytoplasm and Sex

We have mentioned thus far only cases in which ordinary
somatic characters were involved. But there are additional
facts where the possibility of a cytoplasmic influence in sex
determination is under discussion, in both animals and plants.
Correns (1928) studied the behavior of gynodioecious plants like
Satureia and Cirsium, in which pure females occur besides
hermaphrodites, and females pollinated with the pollen of
hermaphrodites produce only females. The same result appears
if the experiment is performed with different species of Cirsium,
one of which is exclusively monoecious; the other, gynodioecious.
The result, viz., only female offspring, was not changed if in
many succeeding generations the females were backcrossed over
again to the hermaphroditic form. From these facts Wettstein
derived the interpretation that the female plants contained
something in their cytoplasm that inhibits the formation of the
male organs. It might be said in favor of such a view that also

CYTOPLASM AND ACTIVATION OF THE GENE 279

in the species crosses of Epilobium, pollen fertility is decreased
by cytoplasmic influence and, further, that a comparable case of
male sterility in flax, studied by Gairdner (1929), and another
in Aquilegia (Skalinska, 1928), found also its only probable
explanation (given by Chittenden and Pellew, 1927) in the
assumption of cytoplasmic inhibition of male differentiation.

In animals, also, the same problem reappears, viz., in Lyman-
tria, but it is much more complicated. The basic fact is that sex
is determined by a quantitative relation between female and
male determining factors. The male factors are genes situated
within the X-chromosomes. The female factor, i.e., the some-
thing that carries sex toward the female side, is inherited purely
maternally, as is proved by the immense body of experiments on
intersexuality. In ordinary crosses, it could be shown that this
maternally inherited female factor is not changed in 10 and more
generations, and the same is also true when long lines of complex
crosses are made, introducing from the paternal side the chromo-
somes of many different sex races over many generations but
leaving the maternal line untouched. The purely maternal
inheritance of the factor controlling femaleness is therefore an
experimental fact, which has stood more decisive tests than
required. Now, in moths the female is the heterogametic sex,
so that maternal inheritance may mean either cytoplasmic
inheritance or inheritance in the Y-chromosome. Certain
genetic facts seemed actually to favor a location within the
Y-chromosome.

Later, however, very elaborate experiments (Goldschmidt,
1934c) proved that actually the property called femaleness is
inherited through the cytoplasm. But this property was found
to differ in different races in a strictly quantitative way. The
differences were called the different types of strength of the female
determiners.

The proof that this property is inherited within the cytoplasm
forces one either to assume cytoplasmic genes, which is improba-
ble, or to attribute to the specific condition of the cytoplasm a
specific influence upon the action of the sex-determining genes.
Goldschmidt (1934c), following a similar suggestion by Dobzhan-
sky, assumed that the cytoplasmic property, which shifts the
balance toward femaleness, is something that controls the level
of the threshold for the action of male determining genes. This

280 PHYSIOLOGICAL GENETICS

assumption makes it possible to bring this cytoplasmic action
upon sex into line with other known cytoplasmic activities.

G. Genes Controlling Cytoplasmic Properties

We have already discussed on different occasions some of the
tacts that show that actually properties of the cytoplasm of
sex cells, especially egg cells, may be controlled by genes. It
may be useful to recount them again at the end of this chapter,
adding also plasmatic properties of the male gametophyte in
plants and chromosomal as well as nuclear behavior, which
itself is probably under control of the cytoplasm (see Boveri,
diminution in Ascaris).

1. The spiral arrangement of spindle fibers in mollusk eggs,
controlling the dextral or sinistral coil in development (see
page 206).

2. All other cases of so-called maternal heredity: Toyama and
Tanaka's work on pigments of the silkworm egg; the work of
Goldschmidt and Katsuki on abnormalities in the reduction
division and subsequent double fertilization in the same egg;
and the others quoted on page 206.

3. Boveri's (1904) work on the control of chromatin diminution
in Ascaris by the cytoplasm may indicate that some of the genes
controlling details of chromosome behavior will actually turn out
to be genes controlling cytoplasmic behavior. The following
cases illustrate this possibility : Lesley and Frost's (1927) gene for
control of chromosome size in Matthiola; Emsweller and Jones's
(1931) gene for control of interstitial localization of chiasmata in
Allium fistulosum; Gowen's (1933) gene producing diploid or
subdiploid eggs in Drosophila; Beadle's (1930) gene for asynapsis,
(1931) gene for supernumerary cell divisions after meiosis, (1932)
genes for sticky chromosomes and pollen sterility.

4. Many genes for pollen characters have been analyzed, e.g.,
the aforementioned waxy gene (Brink), genes for pollen-tube
behavior (Brieger and Mangelsdorf, 1926), and many others.

4. CONCLUSIONS

Thus we conclude that the cytoplasm is mainly the substratum
for genie action, in which all those decisive processes take place
which constitute development and which are steered by the
genes. The specificity of the cytoplasm is therefore one of the

CYTOPLASM AND ACTIVATION OF THE GENE

281

prerequisites of orderly development, and this is tacitly assumed
when the action of the genes is being discussed. The specificity
of the cytoplasm probably increases with the systematic distance
of the forms, and it may find expression in crossing experiments
in a different result of the action of the same genes in different
cytoplasm. Thus far, however, no fact is known that would
force us to assume that specific hereditary traits exist that are
transmitted through the cytoplasm and are individually caused
by a genetic property of the cytoplasm. The plastids of plants
are probably a third independent constituent of the cell in regard
to heredity.

IV. THE NATURE OF THE GENE

The existence of the Wild-type gene, as previously mentioned,

is assumed only on the basis of the mutant genes. Whatever
conclusions upon the nature of the genes are drawn from experi-
mental facts, in the end they must be derived from views regard-
ing the process of mutation. We do not intend to quote every
single utterance regarding the nature of the gene, as almost all
geneticists have expressed their views at some time more or less
completely. We shall report only such theories as have been
derived from a large body of facts and which have been tested
more or less thoroughly.

There are a few a priori considerations regarding the nature
of the gene which have been emphasized in a more or less similar
way by a majority of the workers in this field (e.g., see Hagedoorn,
1911; Troland, 1917; Goldschmidt, 19205; Muller, 1922; Plunkett,
1926; Wright, 1925; Koltzoff, 1928; Mittasch, 1935; Schmalfuss
et al., 1928/f.). A gene is supposed to have the following
characteristics :

1. It is a highly active substance, and it is potent in very small
quantities.

2. The substance of the gene is doubled before each cell
division; it is therefore capable of assimilation and growrth.

3. The gene is able to undergo a definite, sudden, and in many
cases reversible change, called mutation.

4. The mutated gene is perpetuated in the same way, is as
stable as the original gene, up to the moment of another mutation.

The first two points are accounted for best by the assumption
that the gene has the chemical properties of a catalyst (including
enzymes) and more specifically of an autocatalyst, which repro-
duces itself as one of the end products of the catalyzed reaction.
Hagedoorn (1911) was the first to elaborate this idea, which has
since reappeared more or less independently in the writings of the
authors quoted above. This interpretation of the gene, which
defines only its type of action in terms of chemical kinetics, is
independent of all the special ideas that have been elaborated to

282

THE NATURE OF THE GENE 283

account for points 3 and 4 of the foregoing enumeration. At the
present stage of our knowledge, we may assume, then, that a
gene, whatever it is, behaves in the way of an autocatalyst,
catalyzing all those reactions which we have studied in the
preceding chapters.

To account for the facts of gene mutation and the details of
gene action, special assumptions have to be made, the more
important of which will now be analyzed.

1. THE GENE AS AN ACTIVE MOLECULE OR GROUP OF
MOLECULES

As the gene behaves as a unit in heredity, a unit located at a
definite locus in a chromosome, the simplest and most obvious
conception of its nature is the assumption that it is a separate
particle of matter, i.e., one or more molecules of a substance. If
this is the case, a number of possibilities are given for the explana-
tion of the process of mutation. The most important ones, at
the present stage of chemical knowledge, are as follows:

1. A mutation is a change in the quantity of the gene, i.e., from
x to y molecules (Goldschmidt).

2. A mutation is a complete change from one molecule into
quite another.

3. A mutation is a partial change within the molecule, one
residue or side chain being replaced by another one (Guyer,
Correns, Demerec).

4. A mutation is a change within the tridimensional arrange-
ment of a molecule, i.e., the formation of a stereoisomere
(Karczag).

5. A mutation is a quantitative change, not by simple addition
or subtraction of individual molecules but by polymerisation, the
formation of chain molecules (Baur, etc.). We shall review these
possibilities after adding a few remarks concerning the number
of genes.

Authors who assume that the gene is a molecule or a small
group of molecules are naturally interested in proving that the
size of the gene is of the order of magnitude of protein molecules.
Though calculations of this type may not seem so significant
now that it is known that chain molecules exist having a mag-
nitude of 800 to 1000A. length and more (see page 293), the facts
may be briefly mentioned. Estimates of the number of genes

2S »

PHYSIOLOGICAL GENETICS

in a chromosome have been made repeatedly. Morgan and his
colleagues did it in a more general way. Muller (1928a) calcu-
lated from the frequency of unequal crossing over in the Bar
locus. Muller and Prokofieva (1935) made a new calculation;
Bridges drew conclusions from the number of bands in the
salivary chromosomes (ca. 3540); Shapiro and Serebrowskaya
(1934) and Berg (1934) measured mutation rate as compared
with chromosome length excluding the inert regions. The most
elaborate calculations have been made by Gowen (1933), who
used data on the frequency of mutation, both visible and lethal,
after irradiation with different doses and calculated on the basis
of the probability of single hits. Assuming the existence of 175
visible, viable sex-linked genes in Drosophila and finding 7.3
times more recessive lethals in the X-ray experiment, he had to
assume 1280 loci for these genes. Furthermore, 960 loci must be
assumed for dominant lethals on the same basis of estimation.
From the effect upon the autosomes, taking into account their
relative mass, a similar calculation for the autosomes is made,
which leads to large numbers, as Table 21 shows:

Table 21

(From Gowen)

Type of loci

Number

Per cent

Sex-linked loci of visible factors

175
1,280

960

1,800

13 , 100

9,800

175

0.6

recessive lethals

4.7

dominant lethals

3.5

Autosomal loci of visible recessives

recessive lethals

6.6
48.0

dominant lethals

36.0

0 6

If it is taken into account that not all these mutants will appear
at different loci, the minimum count for genes is 1,280 for the
X-chromosome and 13,100 for the autosomes. From these data
the size of the gene may be calculated. The upper limit is set at
1 X 10-18 cm.3 from physical considerations. The actual size, as
computed from the size of the chromosome, would be 10-6 cm.3
This, according to Gowen, would allow for something like 15
protein molecules.

THE NATURE OF THE GENE 285

A. Mutation as Change in Quantity

This interpretation of gone mutation has been derived by
Goldsehmidt (since 1916) from the following facts and delibera-
tions:

1. There is a basic difference of action between one and two
sex-determining genes, showing that the dosage of the gene is
important for its action.

2. Intermediate conditions are found, showing that an orderly-
series of effects exists between the action of one and two sex genes
(intersexuality). In this case, these actions may be described in
terms of intermediate quantities (between 1 and 2) of sex genes.
This enables one to predict the results of any possible combina-
tion or addition of different sex genes, results that were consist-
ently obtained. For this reason, it could be concluded that the
different potencies of sex genes, varying between the potencies
of one or two actual doses, were dosage differences themselves.

3. It was shown that the different quantitative combinations
of sex genes acted by controlling developmental reactions of
definite velocity, which were proportional to the assumed relative
quantities of the sex genes. The simplest assumption that one
can make regarding the action of an active substance, ceteris
paribus, is that it follows the mass laws in which one variable
is the quantity (concentration) of the catalyst. The assumption
that this is the meaning of different gene quantities would there-
fore link the action of the gene to elementary facts of chemical
kinetics.

4. It was found that many cases exist in which the effect of
actually different gene doses is roughly proportional to the dose.

5. It was found that series of conditions of genes exist, the
multiple allelomorphs, in which frequently the effects are pro-
portional to a simple seriation of the genes in question. The
differences of these alleles may be called their potencies; if com-
pared with the other facts, they may be considered actual dosage
differences.

6. It was found that in the cases mentioned under 5, the effect
in heterozygotes and compounds frequently appeared to be con-
sistently equal to the added effects of the respective simplex genes.

From such facts, many details of which have been reported
above, Goldsehmidt concluded that the majority of mutations

I'M) PHYSIOLOGICAL GENETICS

of the gene consist in changes in its quantity, say from 5 to 6 or
to 10 molecules or from ") to 4 or to 2, etc., molecules. This.
again, presupposes that the quantity of a gene (the number of
molecules) is as constant as its quality and that a mutational
change in quantity leads to a new equilibrium.

This theory has been criticized with a number of objections,
many of which are not relevant because based upon misunder-
standings. To mention only one: it has been argued (Muller,
1932a; Timofeeff, 1933a) that the existence of reverse mutations
excludes the quantity idea. It is difficult to imagine what is
actually meant by such an argument. If it is possible that the
typical quantity of gene molecules is controlled by the underlying
chromosome mechanism, and if gene substance is produced as a
result of gene action (autocatalysis), an idea without which no
theory of the gene can work, the number of these molecules may
increase or decrease, whenever the equilibrium is disturbed (this
is the meaning of mutation).

More serious are the following objections:

1. From the standpoint of chemistry. It was pointed out by
Wettstein (1924a) that typical changes in the velocities of
reactions could be produced, e.g., by a change in pH. Schmalfuss
(1928J7.) performed many model experiments with melanin
formation and showed that models can be made up in which
differences in the quality of the chromogen or the oxydase might
lead to different quantitative results in regard to melanin forma-
tion. The correctness of such statements is obvious and coin-
cides, after all, writh everyday experience. If, for example,
growth is controlled by a vitamin and by a hormone, their com-
bined effect is quantitative. Multiple factors, which probably
are different in quality, also combine to make a quantitatively
different action. But such facts have nothing to do with those
cases from which the theory has been derived, in which some
members of the1 series are actual quantities, a fact from which
an inference is extended to the other members. The only
conclusion that may be drawn from criticism of the type here
discussed is that many cases are known in which the gene acts
proportionally to its actual quantity. Also, there are many
other cases in which it is possible to draw an inference from the
knowledge of the action of actual gene dosage upon intermediate
effects of other gene conditions as due to intermediate quantities

THE NATURE OF THE GENE 287

of the gene; but in no other cases can it be proved that muta-
tional changes are of a quantitative nature. This conclusion,
of the agnostic type, is unassailable, as agnosticism usually is,
but it is not constructive.

2. From an evaluation of the facts concerning pleiotropic
action of genes (Dobzhansky, 1930; Muller, 1932; Stern, 1930).
These facts were discussed (see page 89), and it was shown that
there is no reason to assume that very different developmental
processes must be affected in a perfectly parallel plus-minus
seriation, as threshold conditions and numerous interactions
come into play. A perfectly parallel change in regard to different
processes, affected by the same gene and its allelomorphs, is
actually expected to be the rarer occurrence. Wright (1916), who
found the first case of nonparallelism, realized that the explana-
tion has nothing to do with the gene but with the genetic and
other environment in which different developmental reactions
take place (see page 88).

3. The most serious general criticism of the theory discussed
here is that it is very difficult to imagine a physicochemical
mechanism that controls the constancy of the number of gene
molecules in the different loci of a chromosome. Goldschmidt
(1917a) had pointed to some possibilities, but their realization
would require very difficult assumptions indeed. Later, (1932a)
Goldschmidt returned to this problem and pointed out that new
developments of chemistry could supply a better model. To this
point we shall return later (see page 293).

We may draw the following conclusions:

1. Actually genes may act in proportion to their quantity, and
this relation is, ceteris paribus, of the same type as that found
in the kinetics of chemical reactions.

2. In some cases, what has been described as different allelo-
morphic conditions of genes has turned out to be a series of
actual quantities of chromosome material (Bar, deficiencies).

3. In some cases, the facts permit the inference that gene muta-
tions are actually changes in the dosage of the gene.

4. The conclusion that mutations are generally of the type of a
quantitative change in gene substance would have to be supple-
mented by the still missing demonstration of the mechanism to
control constant quantities. Modern developments of the gene

L'ss PHYSIOLOGICAL GENETICS

concepl may, however, Lead to a reconsideration of all these

points, as \\ ill become clear later.

We may point finally to an important point. Since the gene
must needs give rise to two daughter genes of the same constitu-
tion in cell division, it is generally assumed that the gene divides
into two like an organism. In Goldschmidt's theory, this is not
the case, hut the cell produces as a consequence of the auto-
catalytic action of the gene more gene molecules, of which an
equal number is adsorbed to the daughter chromosome at the
time of division of the chromosome. In a recent paper, Haldane
(1935), who is opposed to the theory of gene quantities, comes
to the conclusion that the gene is something specific which cannot
consist of many similar parts because it might be activated by a
single light quantum. Therefore, it cannot reproduce by
fission. From a biological point of view, one might speak of a
parent and two daughter genes. From a chemical point of view,
one must think of a model gene and a copy of it produced in the
cell. For an explanation, Haldane looks forward to wave
mechanics. Similar views have also been expressed by other
authors (Koltzoff, 1935, see page 301).

B. Mutations as Changes of Side Chains or Residues

A set of facts that has led to numerous speculations regarding
the nature of the gene is found in the occurrence of what are
called nowadays unstable genes. The facts were discussed by
De Vries (1903). They were later taken up and analyzed by
Emerson (1914, 1917, 1922) and Correns (1905, 1919, 1928) and
more recently by Anderson and Eyster (1928), Eyster (1924-
1929), Demerec, (1928jf.), Imai (1925/.), et al. Correns (1919)
seems to have been the first to draw inferences concerning the
nature of the gene. A recent review of the literature on unstable
genes has been given by Demerec (19356). The most typical
cases are found in plants where a peculiar variegation is caused by
unstable genes. Generally speaking, they are produced by a
single mutant gene, which must, however, undergo some changes
during individual development, which may be described in
general terms as somatic mutation. It occurs most frequently
as a reversal toward Wild type but also toward intermediate
conditions. The details are rather complicated, as the following
description of the phenotypes by Eyster (1928) shows:

THE NATURE OF THE GENE 289

A continuous series of quantitative variations ranging from apparently-
deep self colors with only occasional color changes that are apparent
through a series of dilute self colors with all gradations of color from
deep red to whitish and with increasing numbers of dots, splashes, lines,
bands and larger segments of darker and lighter colors; and through a
series of variegations varying in pattern from very heavy to extremely
light.

It is obvious that a hereditary return to Wild type is effected
when an area with the return mutation embraces the germ cells.
Only rarely reversible mutations of these unstable genes occur,
but changes to different alleles are found. Many factors seem to
influence rate and type of somatic mutation in these unstable
genes.

Correns (1919) described this instability as a disease of the
gene and makes the following assumptions: The gene consists of a
large molecule to which the same side chain of atoms (same
residue) is attached, say, ten times. This number might undergo
changes in the plus or minus direction under unknown condi-
tions, external to the gene. In the case investigated by him, a
definite ratio of white and green mosaic spots in the plant would
be correlated with the different number of side chains in the
molecule. Though Correns did not develop from this a general
theory of gene mutation, and though he did not actually consider
these unstable genes identical with normally mutating genes (he
actually was opposed to this idea; unstable genes were sick genes
to him), it is obvious that one might conceive of mutation also
as of a quantitative change in the side chains of a gene molecule.
(The facts could of course be as well described in terms of insta-
bility of the number of whole gene molecules (see Goldschmidt,
19286). From the same body of facts Demerec (19316) derived a
related theory of gene mutation, probably also held by many
other geneticists who have not enlarged upon it.

Demerec starts from the rate of mutations. This is normally
rather low, but all transitions lead from rather stable genes to
very unstable ones. In the latter, the rate of change is suffi-
ciently high to be measured, and the effects may be easily
observed in regard to time of occurrence (the coarser or finer
type of variegation), factors affecting the rate, and the end
products. According to Demerec, the end product of the change
is always the same, viz., change from mutant into Wild-type

290 PHYSIOLOGICAL GENETICS

allelomorph. (In ordinary mutation, however, it is the other
way round, and reverse mutations occur, spontaneously or
induced, less frequently.) The change is further favored by the
environment and may occur only in certain tissues, and it is also
favored by the presence of definite genes. These facts induce
Demerec to assume, as Correns did, that the gene is a single
organic molecule, complex, of course, but in general constitution
not different from known larger molecules. We have mentioned
the case studied by Demerec in Drosophila virilis, the unstable
gene for Miniature wing, which forms a multiple allelomorphic
series, two members of which are unstable. No change from one
allele into another occurs; but within an unstable line, changes
in the degree of stability are frequent. From this Demerec
draws the surprising conclusion that changes from one allele
to the other are independent of each other and arise by changes in
different groups of a gene. As so many changes of a gene are
lethal, and as lethals may reasonably be laid to the actual loss or
inactivation of the gene, it might be concluded that only very few
changes in the gene molecule can be tolerated. The more or less
considerable instability of the gene would then mean the presence
of a chemically unstable group. Experiments in treating such
unstable genes with X rays or high temperatures had negative
results (see, however, Gowen and Gay, 1933), from which it is
not concluded that these unstable genes are different from
ordinary mutations but that the effect of treatment is too small
as compared with the normal rate of change to be detected. The
author thinks that the logical conclusion to draw from these
facts, as compared with all the other facts, is that unstable genes
are a different phenomenon from ordinary mutation, as was
held by Correns, who called them sick genes. In a general way,
one might add that conclusions from the cases here reported
upon the normal phenomenon of mutation could be drawn safely
only from facts common to both phenomena and not from those
in which they differ.

There are a number of facts that show that the problem of
unstable genes is far from being solved and that it is quite proba-
ble that a phenomenon sui generis is involved that might lead
to important information regarding the gene and its action. The
cases in question would deserve a much more complete analysis,
especially with the addition of embryological experimentation.

THE NATURE OF THE GENE 291

There is the work of Lilienfeld (1929) on the form of leaves in
Malva parviflora. A dominant mutation produces an inconstant
laciniate type. It is characterized by an inherited tendency to
show highly laciniated leaves in young plants which are gradually
replaced later by more and more normal and even hypernormal
leaves. This, as well as the behavior of heterozygotes, would
appear as a simple problem of morphogenesis which could be
explained easily by genie action upon the rate of production of
something needed for full development of leaves (c/. the vestigial
case). But the types between laciniate and hypernormal
reproduce to a certain extent their kind; the highest grade,
hypernormal, may be kept constant, and it mendelizes with Wild
type. Thus some condition of the gene must be involved. The
details are too different from those in variegation to permit such
a simple explanation as somatic mutation to Wild type. Lilien-
feld herself accepts Correns' (1919) explanation of the sick gene.
Apparently, however, the facts are not sufficiently clear yet,
though they suggest the possibility of furnishing important
information, if analyzed by new methods. This might also be
said of a nearly related case, the "rogues" in peas.

One more case of this type may be mentioned. Anderson-
Kottoe (1931) analyzed a very complicated case of variegation in
ferns, which looks somewhat as if it involved unstable genes.
Visible changes from green to pale tissue occur at the reduction
division, in gametophytic and in sporophytic tissue. As the pro-
thallium is a simple and regularly formed structure, the moment
of these changes may be exactly determined, and the genetic
constitution of a cell may be tested by regenerates from it. The
results do not agree with the simple assumption of somatic
mutations. It is supposed, moreover, that periodic changes
of the gene in question occur, which cannot be of the nature of
losses or gains of parts. Anderson-Kottoe formulates a very
complicated hypothesis. She assumes that a factor necessary
for chlorophyll-formation exists in seven different conditions and
that each condition may be stable or unstable. The latter means
mutation from one state to another at any point in the life cycle
of the plant. Each combination of these states is responsible
for a certain phenotype, stable or unstable and different for the
phases of the life cycle. From the breeding results, a definite
combination for each phenotype is derived, and its mutability as

292 PHYSIOLOGICAL GENETICS

well as the place of it stated. Thus, a very complicated system
isbuill up, bu1 it would be diflicull to state whether or not this is
necessary and whel her or not one could derive from it conclusions
regarding the nature of the gene. A remarkable but not yet
completely analyzed fact has been contributed by Kostoff (1935)
who finds gene instability in form of variegation after species
crosses in Nicotiana.

If we think of the discussions at the end of the last chapter
(Goldschmidt, Haldane) and the facts presented now, it appears
more reasonable to assume that the phenomenon of unstable
genes has to do with the method of "copying" the mother gene,
which may be visualized in quantitative terms of molecules
(Goldschmidt) as well as of side chains (Correns). In this case,
either no conclusion upon the nature of gene mutation may be
drawn, or only one of a quantitative type.

The first alternative would be realized if Stern's (1935)
hypothesis should turn out to be true, viz., that the results
ascribed to unstable genes are actually consequences of somatic
crossovers after certain chromosome rearrangements, involving
also a position effect upon dominance.

C. Mutations as Changes of a Stereoisomers Type

If the gene is a molecule, or a group of identical molecules, a
mutation might possibly occur through the formation of stereo-
isomers. As a matter of fact, some writers actually speak of
mutant genes as stereoisomeres. I do not know, however, of any
geneticist who has tried to derive such a hypothesis from special
facts, and thus far no facts seem to be known that could be better
coordinated by this idea than by any other. A chemical fact
might be mentioned, however, in this connection. According
to Ruzicka, the stereoisomeres of the male sex hormone have no
hormonic effect at all, whereas many of the isomeres of the
female hormone act alike. This one example demonstrates that
it will be difficult to find rules regarding stereoisomeres which
will enable one to describe the genetic facts in simpler terms, i.e.,
to explain them (see also the theory of Karczag, 1928).

D. Mutations as Polymerizations

The idea that mutations might be polymerizations (or depoly-
merizations) of large organic molecules has been used by Baur

THE NATURE OF THE GENE 293

(1924) to interpret series of multiple allelomorphs. This is, of
course, only a modification of the view that the quantity of a
gene is changed by mutation, as such a quantitative change
might as well be an increase in the number of free molecules as a
polymerization. All the arguments applying to one theory
therefore apply also to the other. But some of the recent
developments in chemistry may lead to a reconsideration of these
different types of quantitative views. The work done on chain
molecules (Meyer, Mark, Staudinger) and their micellar con-
glomerations might furnish a new model of the gene, capable of
quantitative changes, as Goldschmidt (1932a, 1934c) and also
Koltzoff (1934) pointed out. No elaboration of such an idea is
yet available. But the same idea may lead in still another direc-
tion, as will be shown below.

2. THE THEORY OF THE GENE AS BASED UPON PHYSICAL
CONSIDERATIONS

The work on production of mutations by X rays, inaugurated
by Muller, has naturally appealed to the physicist, who was
induced to compare the action of X rays in physics with that
of the same agent upon the gene in producing a mutation. The
most elaborate inquiry from this standpoint has been made in
collaboration by the radiologist Zimmer, the theoretical physicist
Delbrueck, and the geneticist Timofeeff (Timofeeff , Zimmer, and
Delbrueck, 1935). The decisive fact for a physical analysis is the
proportionality between X-ray dosis (measured in "r-units")
and rate of mutability. This proportion is a perfectly linear one,
as was proved first by Hanson and Heys (1929). The second
important fact is that the same rule applies to radiation of
different length (7 rays), demonstrating that the effect is inde-
pendent of the wave length. From the genetic side, the impor-
tant points are that mutations are reversible and, further, that
different "genes" react differently upon radiation. Zimmer now
assumes that the biological effect is produced in any case by hits,
whichever of the different theories in this field is used. The facts
of proportionality to dosis, just mentioned, may be expressed
by the equation x = a(l — e~kD), in which x — number of
mutated genes, a = number of irradiated genes, D = dose of
radiation, k — a velocity constant, e = base of natural loga-
rithms. This empirical equation is compared with the general

294 PHYSIOLOGICAL GENETICS

equation covering all cases of effect by a definite dose upon X
areas oul of n possible areas. He finds that the empirical equa-
tion is identical with the general one if the number of hits is one.
From this he concludes t hat one hit suffices to produce a gene
imitation (as bad also been assumed by others). He proceeds to
compare the consequences of the empirical facts with the different
theories of biological action of radiation. The fact that the effect
is independent of the wave length leads him to the conclusion
that the decisive effect is the formation of a pair of ions by the
hit, because this formation is independent of wave length by
definition (ionization = measure of dosage). One excitation,
then, is supposed to be necessary for a gene mutation.

These facts are used by Delbrueck for the construction of a
model. It has to be supposed that the gene is a very stable
molecule, i.e., a combination of atoms with definite positions and
electronic conditions. Such a system might be changed in
different ways, the most important of which is the dissipation of
energy of excitation of an electron. This leads to an ionization in
the neighbourhood, i.e., a rearrangement of the atomic complex by
a single elementary process in the sense of the quantum theory.
A comparison of the energetic consequences of this view with the
facts of mutation furnishes a qualitative agreement of both.

The general conclusion is drawn that mutation consists in a
change of equilibrium of the atoms of a molecule, produced by the
influx of energy from the outside or the oscillations of temperature
energy always present. In detail, this process is assumed to
show parallel behavior to photochemic processes. The primary
process of absorption of a quantum might lead to very different
secondary processes, such as simple steric rearrangements or
dissociation of definite bonds setting free a reactive radical. A
new residue may be attached from the surroundings, and thus
the whole molecule may be changed. From these deliberations
it follows that the gene itself must be represented by such an
indivisible atomic combination or molecule. The authors make
it clear that these conclusions are independent of any assumption
as to whether a gene is a unit or a part of a whole. We shall later
show that the theory of the gene as a unit is no longer tenable.
This will not prevent the application of this theory, especially
the part which assumes that one of the secondary consequences
of the excitation is an atomic dissociation. This theory is then

THE NATURE OF THE GENE 295

a statement in exact physical terms of what was more or less
clearly implied in all other theories of the gene.

3. THE GENE AS AN AGGREGATE OF DIFFERENT PARTS

In opposition to the idea that the gene is a single chemically
active molecule or a group of such, the view has been expressed
that certain groups of facts require for their explanation the
assumption that the gene consists of different separable parts, or
subgenes.

A. The Genomere Hypothesis of Anderson and Eyster

This theory was invented by Eyster and Anderson (1924,
1925, 1928) to account for the facts of variegation produced by
what are now called unstable genes. The facts have already
been reported and analyzed in connection with Demerec's and
Correns' interpretation. Anderson and Eyster assume that
those facts can be explained only if a quantitative segregation of
parts of the gene takes place during development. The gene,
then, is regarded as composed of a constant number of genomeres,
or gene elements, which may or may not be chemically identical.
The usual genetic difference between pigmented and pigmentless
forms is given in the genomeres, all the genomeres C having
mutated into c. If, however, only a certain number of geno-
meres mutate from C to c, an unstable gene with two types of
genomeres is produced. This gene might be divided into its
genomeres in somatic mitoses, producing different combinations
of genomeres up to a complete separation of the two types C and c.

It is obvious that this interpretation is a slightly different way
of describing the behavior of the "sick" gene. It agrees with
Correns and with Goldschmidt in so far as a quantitative element
within the gene is required, which may be conceived as numbers
of side chains (Correns), numbers of molecules (Goldschmidt),
or numbers of genomeres (Anderson). Any of these or similar
conceptions will account for the facts, but also any other explana-
tion of genie changes. It can therefore not be considered that the
facts lead necessarily to the genomere theory.

B. The Hypothesis of the Serebrowsky School

We have reported upon the case of step allelomorphism in the
scute series of allelomorphs in Drosophila. The facts are

296 PHYSIOLOGICAL GENETIC8

presented on page 230, and their interpretation in terms of
development was discussed. The main point was that the
different scute allelomorphs could be arranged in a definite
xrics according to the pattern of bristles that each allele removed.
Each allele acted independently in the compound in such a way
that only those bristles were missing in a compound which both
alleles affected in common. Let ABCD be four different bristles,
and abed their suppression; then the allele scx may produce the
pattern AbcD, and the allele scv the pattern abCD. The com-
pound scx/scv is phenotypically AbcD/abCD and therefore normal
in ACD but with the bristle b missing. If we leave out of this
formula the symbols for the unchanged bristles, we have bc/ab;
i.e., the areas affected are steplike, and the parts common to both
steps (6) are affected. Serebrowsky, Dubinin, and their col-
laborators Gerschenson, Agol, and Lewit (1929-1931) conclude
that this special behavior can be explained only if the steplike
arrangement of the bristle pattern as controlled by the different
alleles has its counterpart in a similar arrangement of parts of the
gene. A projection of the different patterns in the order derived
for each two alleles from the common parts in the stepladder
would then represent a picture of the structure of the gene, i.e., a
structure of a definite length, made up of a series of smaller
elements arranged like steps or as a bundle of sticks the ends of
which do not coincide.

We have already discussed some of the aspects of the case
and mentioned criticism and countercriticism (page 237).
A priori, there is no clear reason why a pattern of phenotypic
effects ought to be projected as a similar pattern into the gene.
Such a projection would help in no way to an understanding,
since it is impossible to see how a pattern in a small part of a
chromosome could control the formation of a parallel pattern
upon the thorax of a fly. If this were imaginable, one ought
to expect that, in general, the organism ought to be a replica of
its chromosome pattern. The old idea of the manikin in the
sperm head would be revived. The idea was therefore rejected
by most geneticists, some of whom tried to find an embryological
explanation for the facts (Sturtevant and Schultz, 1931; Gold-
schmidt, 19316), whereas others assumed that the basic descrip-
tion was erroneous (Child, 1936). Though the originators of the
theory seem to have deserted it now, it ought to be mentioned

THE NATURE OF THE GENE 297

because as keen a thinker as Muller still believes there may be an
clement of truth in it. The newer discoveries regarding the scute
locus, however, shift the discussion to a very different field (see
page 306, position effect).

It should be pointed out also that it is always dangerous to draw
conclusions from exceptional cases without considering whether
the complicated phenotypic behavior is not more likely to be
caused by developmental conditions than by an assumed property
or structure of a gene. Glass (1933a), for example, has described
a case that he likens to the scute case and from which he draws
definite negative conclusions about the gene. He finds in
Drosophila a mutant facet notch which is an allele of facet but
which produces wing notches, without an eye effect. Facet-facet
notch heterozygous is normal, except for 0.26 per cent notched
flies, which resembles the behavior of some of the scute com-
pounds. Glass uses these facts to criticize the quantitative
nature of multiple alleles. But, looking at the case from the
standpoint of development, it is easy to conceive that the hetero-
zygote is actually intermediate when different thresholds for the
two processes (eye and wings) are assumed, an interpretation that
actually goes back to Wright's earliest work.

C. The Hypothesis of Thompson

Thompson (1925, 1931) derived a theory of the gene from the
experimental studies on the Bar locus in Drosophila, especially
those of the Zeleny school. The general idea is that the gene
consists of a main particle anchored in the chromosome, the
protosome, to which a varying number of one or more particles,
the episomes, are attached. Mutation is due to the loss of one or
more episomes and sometimes also to their addition. Two or
more episomes of the same kind are attached to each other and
form a side chain, and different episomes are attached separately
to the protosome. Varying numbers of the same kind of episomes
produce quantitative series of multiple allelomorphs. Qualita-
tive series are based upon numerical variation of more than one
kind of episome.

It is obvious that this theory contains the combined elements
of most of the other theories mentioned. Quantity of active
episomes explains the same facts as quantity of gene molecules,
and the calculations that Thompson makes from the Bar data

298 PHYSIOLOGICAL GENETICS

regarding the number of episomes will therefore lead to results
exactly consistent wit h ( roldschmidt's calculations on the basis of
quantities. For eventual difficulties, the different episomes are
.it hand to be called in when a purely quantitative point of view
is difficult (Infrabar). This theory, which does not seem to have
more heuristic value than the others, has now lost its original
foundatioD since Bar has been proved to be a duplication of a
chromosome segment. To retain the terminology, one would
have to call the whole of the chromosome the protosome.

4. THE GENE AS LOSS OF CHROMOSOME MATERIAL

Serebrowsky (1929) pointed out that gene mutations may not
be changes within a so-called gene but actual losses of chromosome
material. He assumed that chromosomes tend to stick together
and later to break apart again, whereupon small segments are
lost. What appears as gene mutation would actually be a defi-
ciency. In fact, actual deficiencies, lethals, semilethals, and
visible mutations look like a series of different grades of the same
phenomenon, so that with large visible deficiencies at one end,
small deficiencies may be the other end of the series, i.e., recessive
mutations. Thus, assumptions regarding attachment and break-
age, inversions, translocations, and deficiencies are made different
aspects of the same phenomenon, and their phenotypic effects in
all cases are reduced to the action of a deficiency. In this group
would also be included the so-called gene mutations. Serebrowsky
does not commit himself as to how such a small deficiency
( = recessive mutation) is to be understood in terms of genes. A
recent theoretical discussion by Hagedoorn (1934) may also be
interpreted as of a similar type. Stadler (1932) elaborated a
similar idea. He starts from the difficulties in proving that a
mutation is actually a chemical change within a gene. In fact,
many perfectly good mutations have later turned out to be
actual deficiencies. In experiments with X-ray induced muta-
tions, not only the rate of mutation is proportional to X-ray
dosage but also the rate of production of deficiencies. A similar
parallel exists between the reducing influence of dormancy of the
treated cells upon mutation rate and upon rate of chromosome
derangements. A discussion of many facts concerning mutation
in plants shows that all the results may be interpreted in terms of
deficiencies. The obvious difficulties to such a view derived

THE NATURE OF THE GENE 299

from the facts of reverse mutation and multiple allelomorphio
series do not appear insurmountable to Stadler. Regarding the
gene, he thinks that the association of induced mutation with
chromosome breakage does not necessarily exclude the possibility
that mutations are intragenic changes, for it is conceivable that
some change within the gene may be the cause of the break. But
the identification of the induced mutations in general as a class
distinct from the chromosomal abnormalities is not possible in
the experiments with plants. Stadler, like Serebrowsky, does not
draw any conclusions from his interpretation of mutation
concerning the nature of the gene. It might be assumed that he
took it for granted that small deficiencies may be regarded as the
loss of one or a few genes, whatever they are. Serebrowsky, as
we have seen, had definite views about the gene (the step struc-
ture) which, however, were not derived from the evidence
regarding deficiencies.

In a former chapter (see page 176), we discussed Demerec's
work with induced small deficiencies in Drosophila. Of the large
number that he tested, all but one were cell lethal. This excludes
all these deficiencies from being of the same type as recessive
mutations. The latter, then, must be assumed to be still smaller
deficiencies if Stadler's views are correct. We shall return later
to this problem, the significance of which lies in a different
direction.

5. THE GENE AS PART OF A HIGHER UNIT

The ideas regarding the nature of the gene, as thus far reported,
are based upon the assumption that the gene is an independent
unit, a molecule or a group of molecules, attached to the chromo-
some at a definite locus. The chromosome, i.e., its essential
part, would have to be visualized as a string of genes, com-
parable to a string of beads, in which the bead is a unit and the
string a different thing. This does not necessarily include
the concept that these beads, the genes, are completely inde-
pendent of each other. Muller (1932) has occasionally spoken of
intrachromosomal balance, which would mean a certain relation
of interactions of genes within a chromosome as different from
those between genes in different chromosomes. Schmalfuss
(1929) when discussing chemical models of genes pointed out
that the relative position of gene molecules might have an effect

300 PHYSIOLOGICAL GENETICS

upon their action without any other change. His model is two
molecules with side chains of opposite activity, which when very
near together mighl neutralize each other though not if distant.
This would be a model for what is called now the position effect,
which also indicates that individual genes are not independent
from one another. ( Details will be discussed later.)

A. The Gene as a Side Chain of a Molecule

Actually, theories of the gene have been conceived thai regard
the gene only as a part of a whole. Repeatedly geneticists have
declared that they consider the chromosome itself as a giant
molecule, the side chains of which represent the genes (Castle,
1919; Renner, 1920). Koltzoff (1928) has elaborated this idea.
To him (as to many geneticists) the chromatic part of the
chromosome means nothing but a substratum which surrounds
the essential part of the chromosome, a chromatic thread repre-
senting the gene string of other authors. This is either a very
long protein molecule or a bundle of such molecules around which
new molecules are assimilated from the surrounding chromatin
mixture so that the bundles grow and may divide. In these long
molecules, the many different radicals have their definite position,
and all chemical changes in these radicals, e.g., removal of one
or the other atom or replacement by another residue, will appear
as a mutation.

It is obvious that this view does not make it easier to under-
stand genie action. Its importance is that it constitutes a first
step toward a definition of the gene as a part and not as an
independent unit.

Koltzoff (1934, 1935) has repeatedly returned to his ideas and
modernized them. He compares (page 293) the recent develop-
ments in chemistry with his views and brings them in line. The
chromonema or gene string or genonema is a chain molecule (or
micella composed of such molecules) containing protein and other
radicals, the genes. From the surrounding solution these radicals
attract their like and synthesize them, assimilate them. The
surplus of assimilated molecules, i.e., the genes, their parts, or
complexes, pass into the cytoplasm, and thus the genes produce
their effects. Complicated and necessary substances like sterols
and agglutinins are also supposed to be actually present as genes,
which assimilate their like. It may be added that more specific

THE NATURE OF THE GENE 301

views regarding the genes themselves are not expressed in these
papers.

Apparently, Goldschmidt (1930) was the first to point to cer-
tain recent developments in chemistry which could lead further,
as did also Koltzoff (1934, 1935) as mentioned. Quoting the
work on the chemical structure of fibers (Meyer and Mark) which
showed that large chain molecules of measurable length combined
into a micellar bundle make up these structures, he pointed
out that genes may turn out not to be free molecules but mole-
cules combined into a micellar structure. He alluded to the
possibility that the "quantity of the gene" might thus turn out to
be the number of links in a chain molecule. In another paper
(1932a), he says: "The latest developments of organic chemistry
suggest a different relation which is also of a quantitative type
and nevertheless capable of being produced by the addition of a
single quantum: changes in the length of members of a chain
molecule in both directions. I shall not develop further this
idea, to which I think the future belongs."

The same idea has recently been taken up by a chemist Miss
Wrinch, who has developed from it a theory of the gene (1934,
1936), going into the chemical details of what, with Goldschmidt,
was only a suggestion and with Koltzoff, little more.

The starting point for Miss Wrinch's speculation is identical
with the viewpoint of all geneticists who have tried to advance a
theory of the gene. She writes (1936):

The regions of the chromosome of cytology associated with certain
genes have become ever smaller and smaller, in some cases being no
longer than a hundred angstroms. Simultaneously molecules of the
chemist have become ever longer and longer, now reaching a length
greatly exceeding 100 angstroms. There is no resting place for the
theorist concerned with the structure of chromosomes until his postulates
are capable of expression in molecule terms. In a dark forest of facts
this at least stands out as clear as day — only with the help of molecular
physics is it possible to devise a structure for the chromosome.

The chromosome, then, is a molecular aggregate, and the
problem is the chemical nature of the constituent molecules
and the manner in which they are arrayed. Wrinch starts
with the protein molecule and stresses the fact that the basic
molecule is built largely of amino-acid residues and that the
active proteins have a high content of diamino acids, especially

302 PHYSIOLOGICAL GENETICS

arginine. As a model, she takes the clupeine. It is said to

consist of Ions chain molecules, in which arginine residues
alternate with certain nionoaminoacid residues, 4 or 5 of the
former to 2 of the latter, altogether 28 residues, with the length
of the chain 98A. Such molecules, placed end to end and held
together by suitable bonds in a chain, form the protein pattern
of the chromosome, and a bundle of these constitutes a chromo-
some micella. The individual molecules are different poly-
peptids, and the pattern may therefore be expressed in terms of
side chains of the constituent molecule. The bundle, the
micella, may have any width without changing this pattern.
Any difference in this pattern, be it of one residue (estimated
length 3.5A.) or of one molecule (98A.), would constitute a
genetical difference. Wrinch points here to the position effect,
which would fit into such a picture. A calculation shows that
such a pattern of chromosome length may vary to the number of
2Q5o,ooo patterns. Wrinch then proceeds to find the proper place
for the nucleic acid, also present in the chromosome. She views
the constitution of this molecule as possessing four negatively
ionized centers, which are bound to form basic compounds with
the positive amino acid residues. If all four centers do this, and
if, as assumed above, all the chains within the micella are identi-
cal, the nucleic acid ions will completely encircle the micella.
This is the reason why the protein molecules aggregate : they are
held together by these nucleic acid ions. This relation may also
be expressed by comparing the protein pattern to the warp
and the nucleic acid to the woof of a fabric. It may be added
that if the nucleic acid molecule is highly polymerized, it may
hold together, instead of four, a large number of parallel protein
molecules.

In harmony with this concept of the chromosome, Wrinch
assumes that a gene is a group of molecules between the natural
breakage points, viz., the bonds between the individual chains of
molecules. We shall return later to this hypothesis which may
also be adapted to a very different point of view.

B. Friesen's Hypothesis of Chain Mutation

In an earlier chapter we mentioned the hypothesis of Dubinin
(page 296) that the gene is composed of subgenes, or centers,
which mutate independently; the mutated gene is the sum of

THE NATURE OF THE GENE 303

the changes of the centers, which are arranged after the model
of steps. The facts were given upon which this concept was
based, namely, the behavior of the scute alleles. But it is
known that other series of multiple alleles do not exhibit compar-
able phenotypic effects. In the chapter on multiple allelomorphs,
we considered those cases in which manifold effects of multiple
alleles seemed to behave rather independently of each other.
Friesen (1931), who analyzed one such ease, thinks that they must
be explained (also the scute case) by the assumption that a single
gene never mutates but always a more or less long chain of genes.
He speaks of chain mutations, which he believes to be the
prototype of mutations. This idea, which it would be difficult
to prove, is in principle not different from the subgene concept.
In both hypotheses, actual or apparent difficulties in explaining
phenotypic effects as due to a single gene-controlled process are
surmounted by projecting a corresponding multiplicity into the
germ plasm. The subgene theory does it by subdividing the
gene; the hypergene theory, as we might call it, does the same by
assuming a change in a chain of genes. The same criticism (see
page 296) applies therefore to both conceptions. The theory
has, however, the significance of showing the growing need of
considering the gene only as a part and not as a unit. Friesen
actually says that the assumption that genes are united into
groups means that they are not independent physicochemical
structures but that, in spite of a certain independence, they show
some structural interrelations. The same statement may also
be found in Brink's work (1932). He concludes from the facts
relating to translocations in maize that the chromosome consists
of groups of physiologically interdependent genes. Therefore,
the propinquity of the genes within a group is essential to normal
gene action in a general physiological sense.

6. OUTLOOK UPON THE THEORY OF THE GENE OF TOMORROW"

We have mentioned a number of facts that point in the direc-
tion of a conception of the gene more as a part of a higher unit
than as an independent one. The developments in the last years
of genetical research have now led to a point where it has become
necessary to ask the radical question : Is not the whole conception
of the gene as a hereditary unit obsolete? The facts that make

304

PHYSIOLOGICAL GENKTKS

such an inquiry necessary arc grouped around a phenomenon

that -rather unfortunately — has been called position effect.

A. The Position Effect

This phenomenon was discovered by Sturtevant (1925). We
have reported in different chapters the details of the Bar-eye case
in Drosophila. Bar is a duplication of a small segment in the
X-chromosome. By unequal crossing over, two Bar genes (if we
continue using this term for descriptive purposes) may come
together in one chromosome; when homozygous, this is called
Double-bar or Ultrabar. At the same locus there is a mutation
Infrabar, which also may be obtained by unequal crossing over as
Double-infrabar and together with Bar as Bar-infrabar. We
have analyzed the quantitative effects of all these combinations
expressed in facet numbers and found them proportional to the
number of genes present. A combination B/B and a combination
BB/ have both two Bar genes and ought to have the same
effect. Sturtevant found it to be different (see Table 22).

Table 22
(From Dobzhansky)

Genotype facets

Genotype facets

BB/
B<B</
BB</

45 . 42 ± 0 . 24
200.2 ± 8.6
50.46 ±0.40

B/B
B'/B*
B/B'

68.12 ± 1.09
348.4 ± 12.4
73.53 ± 1.29

This means that two Bar genes in the same chromosome
prevent facet formation in a higher degree than do the same
genes in different chromosomes. This might also be expressed
thus: Two adjacent Bar genes reinforce each other's action, as
compared with two opposite Bar genes. The position of the
gene has an influence upon its action.

This effect came into the foreground of genetical interest when
chromosome breaks, translocations, inversions, and such rear-
rangements of chromatin material were obtained by X-raying
and studied extensively. Other cases of this effect came to light,
and the same term was applied to a number of related phenomena.
We may arrange the facts in groups which are not very natural
and are only provisional, as new facts are frequently being
discovered. Very probably the different groups will turn out to

THE N ATI RE OF THE GENE 305

be only variations of the same phenomenon. All the work was
done with Drosophila.

1. A first type may be described as follows. At the locus of a
known mutation, a break in the chromosome is produced, and a
translocation between this chromosome and another chromosome
occurs. The new chromosome arrangement is accompanied by a
phenotypic effect, described in ordinary genetic language as a muta-
tion, wrhich behaves as does an allele to a known mutation at the
same locus with the same type of phenotypic effect. Dobzhansky
(1932) analyzed such a translocation. The X-chromosome was
broken at the Bar locus, and the left end was exchanged for the
right end of the second chromosome. The result was an effect
of the type of a Bar allele and was called Baroid. In this case, it
could be fairly well proved that nothing else had happened
and that the Baroid phenotype is a result of a new arrangement
at the Bar locus, i.e., a position effect. (Here the term has a
different meaning from the one in the original Bar case : there it
described the actual facts; here it implies a definite interpretation.
But as the term is widely accepted, it may as well serve for
descriptive purposes.)

Since Dobzhansky's discovery was made, other such breaks
at the Bar locus have been reported with a corresponding effect.
In some cases, it is stated that the break is only near the Bar locus
(Offermann 1935). We shall return to this point later. Dubinin
and Volotov (1935) produced numerous breaks at the Bar locus
by X-raying, many of which — but not all — produced the Bar
effect ; and Dubinin (1936) described another case of Bar mutation
similar to Baroid, produced by a translocation between chromo-
somes I and II writh the break at the Bar locus. Altogether they
produced 34 mutations of the Bar type of which 10 proved to be
breaks at the Bar locus, 1 a deficiency at that locus, and 6 near by.
In the case of 7 incomplete reversions thus produced, 5 had a
break at some distance from Bar.

2. The second type of position effect is of the same order as the
first; i.e., a break at a definite locus occurs, and the phenotypic
effect of an allele of this locus is produced. But in this group the
rearrangement takes place within the same chromosome, usually
as an inversion. It had been known for a long time that some
of the alleles of the scute series were accompanied by chromosome
breaks, also alleles of yellow and a few others (Serebrowsky and

306 PHYSIOLOGICAL OENETIC8

Dubinin and (heir collaborators; Muller, Patterson, and Stone;
Beadle and Sturtevant et ah). The work of the last-named
group of authors showed that breaks at or near the scute locus
were involved. The more refined study of such cases by the
salivary-gland method revealed that actually very small inver-
sions were present. Muller and Prokofieva (1934, 1935),
Muller, Prokofieva, and Raff el (1935) made such an analysis
for the left end of the X-chromosome. They showed that
different rearrangements involving the scute locus in Drosophila,
in the great majority of cases, result in phenotypically different
"allelomorphs," whereas nearly identical rearrangements (scute
4 and scute 68) have given essentially the same allelomorphs.
Of 27 scute and achaete mutations, 18 were proved to be caused
by rearrangements after breaks, and the others are suspected of
being the same. One case (scute 19) is a deletion of only a single
segment (in the salivary chromosome) with insertion in chromo-
some 2. Other scutes proved to be inversions of small fragments
in situ. In this and similar cases, two effects had been observed
(double mutations) which thus are referred to the two points of
breakage of an inversion. From such facts the authors are
inclined to assume that possibly all mutations produced by
X rays may be such minute rearrangements.

But such effects are not necessarily bound to minute inversions.
Griineberg (1936) reported upon a very long inversion within the
X-chromosome which behaved like an allele of roughest and
produced a very rough eye surface. He wras able later to get a
return arrangement to a normal order in the chromosome, and
with it the effect disappeared. What seemed to be a mutation,
then, was a position affect.

3. A third type of position effect was described by Dubinin and
Sidorov (1934, 1935) for translocations between the fourth and
other chromosomes. In this case, the effect of the break does not
resemble a gene mutation but a change in dominance, wrhich, of
course, could also be described as the production of a mutant
dominigene. In flies heterozygous both for this translocation
and for the fourth-chromosome recessive cubitus interruptus, this
latter character appeared as semidominant to completely domi-
nant. Sturtevant and Dobzhansky (Dobzhansky, 1936) tested
some more such translocations, so that altogether 48 have been
tested, of which 22 showed the dominance effect. Dubinin

THE NATURE OF THE GENE 307

performed also the necessary tests to exclude other explanations.
He found, further, a similar result for the gene hairy in the third
chromosome — in fact, both effects simultaneously, when the
translocation involved breaks near both the hairy and the
cubitus interruptus loci. In a crossover experiment, Dubinin
even claims to have introduced the hairy locus itself into the
broken partner chromosome with the result that this replaced
hairy gene also showed the dominance effect.

Similar results were obtained by Dobzhansky and Dobzhansky
(1933) for the gene bobbed and by Dobzhansky and Sturtevant
(1932) for the genes yellow, kurz, rudimentary, and forked. But
in these cases duplications were involved, i.e., chromosome
fragments containing the Wild-type alleles in addition to the two
recessive genes in the normal chromosomes (one + being domi-
nant over two recessives). The increased dominance of the
recessives in these cases might have different reasons, usually
expressed in terms of disturbance of genie balance. But the
remarkable fact is that the dominance-shifting effect is produced
only when the duplicate fragment is broken off near the locus
of the gene showing the dominance effect. This is certainly
in favor of the assumption that here, also, the position effect
of the cubitus interruptus type is involved.

4. There are cases that belong to more than one of these
categories; i.e., allelomorphic mutations are found as a conse-
quence of translocations, inversions, duplications, and deficiencies.
But the decisive point is that, again, the effective locus is near or
identical with the point of breakage. A complication within this
group of cases is that these effects behave frequently like unstable
genes and therefore appear as variegation, mottling. All these
cases are phenotypically dominant eye colors, studied first in
detail by Muller (1930) and later especially by Glass (1933, 1934)
and Schultz and Dobzhansky (1934). Without going into the
complicated and not yet completely analyzed genetical details
(now checked by the salivary-chromosome method), it may be
stated that in the best known set of Plum alleles a break has
always occurred near the brown locus in the second chromosome
to which all these mutations are allelomorphic and another one
near the light locus. According to Schultz (1934), in these
cases a transfer of chromosome material to the inert region
(chromocenter in the salivary chromosomes) is always involved,

308 1'IIYSIOLOGICAL GENETICS

and other authors have confirmed this result. In general terms,
we are dealing with the same type of position effect at or near a
break, but there is the additional fact that the Plum effect appears
only when the material from the brown locus comes in touch
with the chromocenter.

5. In a further study of the last-mentioned Plum rearrange-
ments, Dubinin (1936) found what he considers to be a new type
of position effect. By further breaks in the inverted region of
the second chromosome, which is responsible for the Plum effect
by bringing a section of inert material in contact with the broken
chromosome, he produced chromosomes in which both ends of the
inversion are in contact with inert material. He found that
considerable regions on both ends of the inversion produce the
Plum effect if contiguous with inert material. The position
effect is therefore not a point effect in this case but an effect of
different regions. This result ought to be considered in connec-
tion with the frequently found position effect not at but near the
break.

B. Interpretation in Terms of Genes

Different interpretations have been proposed to explain such
facts without changing the concept of the gene. We mention
only the idea that in all these cases a mutation arises near or
at the locus of breakage, which has been disproved in some of the
instances reported; or the idea that small deficiencies are caused
at the points of breakage, which also does not agree with some
of the facts. The position effect as such has therefore to be
accepted, and the problem is, to determine what it actually
means. Dobzhansky (1936) expresses himself in a more general
way:

The activities of a gene lying in an abnormal position in the chromo-
some are deflected from their normal course by influences emanating
from its new environment but no permanent alteration is wrought in the
gene itself. . . . Position effects may prove to be essentially develop-
mental phenomena, but phenomena of intracellular physiology, affecting
the coordinates on which the reactions between the primary gene prod-
ucts take place. . . . The hereditary material is discontinuous, for it
is segregated into independent units, genes. And yet, it is a continuum
of a higher order; since the independence of the units is incomplete, they
are changed if their position in the system is altered.

THE NATURE OF THE GENE 309

More specific arc Muller and Prokofieva (1935) when they
assume that position effects are due to a higher degree of inter-
action between locally more concentrated products of gene
activity as compared with the interacting of products from widely
separated points, which are either more diluted or changed (see
Schmalfuss, page 299). The same idea is also offered by Offer-
mann (1935) and is found in Sturtevant's first discussion (1925).
Muller (1935) will not exclude absolutely the possibility of inter-
action between the genes but thinks that it is more probable
that only the products of the genes are involved. In a lecture
before the Physiological Congress of 1935, he elaborated this idea
as follows. In the production of phenotypic effects, the gene
begins by interacting with cellular substances to produce specific
products which diffuse from the locus of origin and cause or affect
further physicochemical changes. Then follow the chains of
reactions and interactions, conceived in the same way as we
discussed them above. When such interactions involve the
immediate products of neighboring genes there, distance may
count. It is to be assumed that the highest concentration of the
gradient of gene-controlled primary products will be near the
gene, whereas with increasing distance these substances will be
altered by undergoing new reactions. This idea has been further
elaborated by Offermann (1935).

The important point for the discussions of this chapter is that
it is generally assumed that the position effect has to be explained
within the present gene theory, which remains unaffected as such.

Muller and collaborators (1935) once were tempted to go one
step further, when they argued: "As our studies of mutations in
the X and other chromosomes have shown that apparent replicas
of practically all known natural mutations in Drosophila may also
be obtained by X-rays, the further question is raised as to what
proportion of natural mutations in Drosophila may really be
minute rearrangements." The importance of trying to dis-
tinguish intragenic from intergenic changes is therefore urged.

C. Is the Theory of the Gene Still Valid?

The preceding sentences bring us now to the point where we
have to ask ourselves whether or not the theory of the gene as the
hereditary unit of actual separate existence is still tenable. The
facts regarding position effects, which we mentioned, have led to

310 PHYSIOLOGICAL GENETICS

a situatioD where genelike effects are attributed to contiguity
between different points in a region <>f the chromosome, assumed

to represent different genes, and the so-called inert material (see
the Plum case, above). The theory of the gene has certainly to
be stretched considerably to allow a description of such facts in

terms of genes. Is there no alternative? It seems that these
facts and a number of others, to be mentioned, point to a theory
of the germ plasm in which the individual genes as separate units
will no longer exist. The gene as a unit is, of course, a concept
derived from the existence of a thing called the mutant gene.
The normal condition allelomorphic to the mutant condition is
assumed to he controlled by the plus gene, when it can be proved
that the mutant effect is localized at a certain point in the
chromosome, called the mutant gene. But this inference is not
necessarily valid. There is a possibility that a condition exists
at a definite locus of a chromosome that we call a mutant gene
but that no corresponding plus condition exists as a separate
unit. Let us assume, with some geneticists (to whom the
material was not yet available for a serious discussion of the
point), that the whole chromosome is a large chain molecule of
complex arrangement. Each point in the chain, i.e., the residue
attached at each point, has a definite meaning in the chemical
properties and the reactivity of the wdiole, as is the case with all
molecules. (For a detailed model, see Wrinch, page 302 and
later discussions of the work of Bergmann.) The Wild type,
then, might be controlled by the whole chain as a unit. If the
chain is intact, and each residue in its proper steric position, the
catalytic processes dependent upon this chain molecule occur
in a way that leads to what is called a Wild type. Any change
in the chain, however, of wdiatever type, may disturb the normal
interplay of the catalyzed reactions, and a deviation occurs
which is called a mutation. These changes described in terms
of chemical models may be of several types, such as the following:
(1) the differences of stereoisomeres, which means that each
inversion or other rearrangement within a chromosome would be
equivalent to the production of a steriosomere; (2) the differences
in the length of chains as they characterize for example, the
different carbohydrates. Each breakage of a chromosome would
therefore result in a change of the chemical properties of the
molecule; (3) the possibilities of chain molecules composed of

THE NATURE OF THE GENE 311

unequal links, the manifold combinations of which form related
but different compounds. Translocation would lead to such
differences. In general, if the chain molecule (= chromosome as
a unit) is needed to produce the reactions controlling the Wild
type, any deviation from the typical steric arrangement would
change the chemical properties and produce a different type, the
mutant, without any change of residues or atoms. As this
steric change happens somewhere in the chain, the mutant type is,
of course, related to a condition at a definite point or more than
one point. Such a point we call the mutant gene. Let us assume
that the order in the chain has been changed by an inversion:
the new effect appears to be localized at one or both points of
breakage, and we might use the word mutant gene for these
points, though there is nothing special situated at this point
that we could describe in terms of matter, the point being actually
a point of reversed order of the old constituents. If, however,
no change in the order occurred, we have no possibility of describ-
ing this point as a unit of action and therefore cannot call it the
Wild-type gene. In other words, there is such a thing as a
mutant gene, if we choose this expression for a stoichiometric
change at a definite point in a chain molecule, which has the
effect of a change in chemical properties, whatever the change
may be. But there is no Wild-type allelomorph, only one single
normal arrangement of the chain molecule as opposed to all
other possible arrangements. The facts of genetics may, of
course, be described in terms of genes, but a theory of the germ
plasm would have to do away completely with the concept of
genes as units. We have intentionally not mentioned the
possibility of chemical changes in the individual residues, because
the facts, which force us to revise the conception of the gene, are
concerned only with steric and stoichiometric changes. Whether
or not chemical changes in the residues occur or what may
be their role (probably in phylogeny) cannot be derived from the
genetic facts thus far known.

The following facts seem to point in the direction of a view
regarding the nature of the germ plasm and of mutations of the
type as just described:

1. The action of X rays upon chromosomes produces pre-
dominantly breaks and the different known rearrangements ; the
same treatment produces the so-called gene mutations.

312 PHYSIOLOGICAL GENETICS

2. The proportional relation between the quantitative effect
of X rays ami the amount of ionization is parallel for mutations
and rearrangements.

3. Temperal lire shocks that increase the rate of mutal ions also
produce chromatin rearrangements, in both animals and plants.

4. Mutations induced by X rays and spontaneous mutations
are identical.

5. The phenotypic effect of rearrangements, i.e., the so-called
position effect, under which must be included all the effects that
have been described as mutations occurring simultaneously with
rearrangements, is of the same type as the effect of so-called
gene mutations. We find all the known types of genie effect.
In fact, many effects that originally were regarded as due to gene
mutation have turned out to be position effects of rearrangements.
Specifically, the following types may be mentioned, using the term
position effect now for all phenotypic consequences of whatever
rearrangements.

a. Dominant and recessive effects.

b. Effects of the type of a dominigene effect (cubitus inter-
ruptus) .

c. Different position effects produced by different rearrange-
ments involving the same locus behave like a series of multiple
alleles (scute, notch, Bar).

d. "Invisibles" — mutations or rearrangements without visible
effect.

e. Effects of the modifier type; e.g., the so-called mutant
Beaded produces alone only a nick in the wing of a percentage
of individuals. If combined with an inversion, this effect is
modified into the real beaded type. The inversion may be
replaced by others in the same region with the same effect, and
the removal of the inversion also removes the modifying effect
(Goldschmidt, unpublished).

/. Cases in which the effect of the mutant gene is proportional
to its dosage. There are chromosome rearrangements exhibiting
the same phenomenon, e.g., the Bar case.

g. There is a most remarkable parallel between the spontaneous
appearance of so-called gene mutations and effects of rearrange-
ments. Goldschmidt (in press) analyzed a series of cases of
complicated rearrangements occurring spontaneously within
different lines of Drosophila, which led to the appearance of a

THE NATURE OF THE GENE 313

number of phenotypes. Some of these behave genetically like
ordinary new mutants, a considerable number of others being
identical with known standard mutants. If this had not occurred
in pedigreed and controlled lines but in an ordinary stock,
the impression of a series of simple gene mutations would have
been created. Actually a similar case found by Plough and
Holthausen (1937) has been interpreted as mass mutation of
genes. There are cases in which a number of well-known mutants
occurred in a way that is rather suggestive of a similar origin.
For example, in Drosophila, truncate originated in a Beaded
stock; vestigial and balloon, from truncate; spread, from Beaded
and vestigial; purple, from vestigial; kidney, from vestigial X
purple; ebony, from balloon. Blistered, jaunty, and curved
came from rudimentary (two of which we found to be produced
by a rearrangement). Many new mutants appeared among the
progeny of crosses between other mutants.

h. It is most remarkable that rearrangements within certain
areas have a similar phenotypic effect and might therefore be
described as different multiple allelomorphs, e.g., a number of
deficiencies near the locus plexus produce so-called plexates.
Probably some translocations in this region have a similar effect.
Duplications and translocations in the Bar region produce the
Bar effect; different deficiencies in the fused region, notch effects;
point mutations and translocations in the yellow region, yellow
color. The similar phenotypic effect of different happenings
at one point or in a more or less long chromosome section near
this point may in these cases be due to a so-called mutation, a
deficiency, a translocation, a duplication.

i. There are cases in which the end members of a series of
multiple allelomorphs are suspected or known to be deficiencies
(vestigial, eyeless). There are other series of multiple allelo-
morphs with an additional deficiency appearing as a member of
the series, e.g., bobbed.

j. There are certain changes that occur at many different
points of different chromosomes, producing a similar phenotypic
effect. The minutes in Drosophila are such changes, which in
part are certainly deficiencies and in part not visibly different
from gene mutations.

k. There are a few extremely typical "mutants" that we have
discussed in connection with the cases of homoeosis. They are

::i I PHYSIOLOGICAL GENETICS

all located in the same section of chromosome III: bithorax,
58.7; bithoraxoid, .")'.». ."j; aristapedia, 58.5; proboscipedia, 45.7;
tetraptera, 51.34(?).

Though not complete, this list of facts seems to be a formidable
(me All points taken together suggest strongly that the
chromosome is the actual hereditary unit controlling the develop-
ment of the Wild type that purely steric changes at the individual
point- of its length produce deviations from Wild type which
may he described as imitations, even as point mutations, though
no actual Wild-type allelomorph and therefore no gene exists.

If such a view, which we consider to be tomorrow's theory of
the germ plasm, should turn out to be unavoidable, a number
of problems will have to be solved. The first one is whether or
not a model of such a chain molecule is available and whether or
not it is possible that the different points in its length exercise a
different specific, and independent catalytic action. It seems
that the most recent work of Bergmann and coll. (1937) on the
structure of the protein molecule has furnished such a model
(Goldschmidt, in press). It will be found to be nearly related
to the other models, especially that of Wrinch, with the advantage
of not using the conception of the gene any more.

Bergmann starts from certain facts upon which he builds a
chemical hypothesis, and both facts as well as hypothesis appear
very helpful for our present purposes. The protein molecule is
known to consist of a chain of amino acid residues linked by
peptide bonds. According to Bergmann these residues are
arranged in a simple order. Each amino acid has its own rhythm
different from that of the others, and the superposition of all
these rhythms produces the pattern of the molecule. The
rhythm itself, meaning that one type of residue appears always
separated by a definite number of others in the chain, follows a
simple arithmetical rule. Thus silk fibroin which contains
glycine, alanine and tyrosine residues shows an arrangement of
these in the rhythm:

G-A-G-X-G-A-G-X-G-A-G-T-G-A-G-X-G-A-G-X-G-A-G-X-
G-A-G-T-G (G-G = 2, A-A = 4, T-T = 16)

The total number of members of a chain is 288 or a multiple
thereof. To organize such a molecule, a chemical organizer is
needed of immense specificity. According to Bergmann the

THE NATURE OF THE GENE 315

only known substances with the necessary properties are the
intracellular proteinases, which both hyclrolyze and synthesize
the protein molecule. At this point Bergmann pronounces the
hypothesis that the proteinases are themselves proteins. "If
the proteinases themselves are proteins and at the same time
have the ability to synthesize other individual proteins then
there must exist proteinases which have the ability to synthesize
replicas of their own structural pattern and therefore are able to
multiply in suitable surroundings."

It is astounding to see how these chemical facts and hypotheses
fit the requirements for a chemical theory of heredity, as postu-
lated in our previous discussion. Let us assume that the indi-
vidual chromosome actually is a single immense chain molecule
and a proteinase. (This means the essential part of the chromo-
some— the so-called gene string — to which is added nucleic acid,
which makes in some way the stability of the long chain possible.
All modern hypotheses regarding chromosome structure have
reckoned with these two constituents.

This proteinase then has different effects according to its
special surroundings. Either it may produce its own replica,
which amounts to a division of the chromosome, or it may
synthesize other proteins from the parts present, or it may
hydrolyze proteins. The latter two activities would constitute
what is usually conceived as being the function of the gene.
The immense specificity of this proteinase is based upon its
typical structure. This, however, represents a most complicated
though regular pattern, composed of the superimposed rhythms
of the different amino acid residues. This again means that any
breakage or rearrangement in the chain leads to a destruction
or impairing of the specificity and therefore to other reaction
products, which have to be assumed to control the phenotypic
changes. Thus it seems that the latest developments of genetics
and of protein chemistry permit the statement of a reasonable
hypothesis regarding the chromosome, which could easily be
elaborated much further.

Among the facts which will have to be explained on the basis
of this new model there is the question how the relations between
quantities of what is called a gene and a proportional effect can
be explained in terms of steric changes. There is no difficulty
in the cases where duplications or deficiencies are involved, as

316 PHYSIOLOGICAL GENETICS

models arc available for a proportional chemical action with
different degrees of methylation, etc. As a matter of fact, the
concept of the gene, which considers its primary property to be
its quantity, will be most easily incorporated into the new hypo-
thesis and will turn out to have been a better guess than most of
t he o1 her concepts (see page 293). This docs not mean, however,
that such physical considerations as those discussed above
(Timofeeff, Zimmer, and Delbrueck) are not valid. They actu-
al ly lead only to the conclusion that the elementary quantum
process produces secondary changes of atomic equilibrium.
One of these is also a dissociation of bonds, which, after all,
would also be needed for the breakage of a chain molecule.
There is, then, no obstacle from the side of physics to changing
the concept of the gene in the direction here espoused. As a
matter of fact, I had reached this conclusion years before the
physical analysis of the three authors was published (see
quotation, page 301).

A second point concerns the problems of how evolutionary
changes occur and whether or not chemical changes in the residues
as opposed to steric changes in the chain molecule are of primary
importance. No facts based on experimentation are yet known
that permit conclusions on this point. But it would be a mistake
to reject necessary conclusions concerning the gene concept,
because they will not lead to an explanation of everything.

Views regarding the gene as developed in this chapter have been
"in the air" for some time. Goldschmidt's remarks (1930a)
have been quoted, and it may be added that he has discussed the
contents of this chapter repeatedly in public during the past years.
The Drosophila workers, while discussing the position effect,
touched upon it; and Muller (1935) almost drew such conclusions
though shrinking, as it seems, from the last step, namely the
abolition of the gene concept. The chemical models used by
Koltzoff and the more modern one by Wrinch are also steps
toward the views developed here. It remains for the physico-
chemist to decide whether or not the new model could also take
care of the independent catalytic actions of what was considered
to be a gene.

BIBLIOGRAPHY

Abderhalden, E. (1936). Neue Wege zur Erforschung spezifischer
Strukturen in Eiweissstoffen. Wien. med. Wschr. 1.

Agol, I. J. (1931). Step allelomorphism in Drosophila melanogaster.
Genetics 16, 254-266.

. (1932). Das Sichtbarmachen der verborgenen allelomorphen

scute-Teile mit Hilfe von Faktorenausf alien (deficiencies). Biol.
Zentralbl. 52, 349-367.

Anderson, E., and L. B. Abbe. (1933). A comparative anatomical study
of a mutant Aquilegia. Amer. Naturalist 67.

and D. De Winton. (1935). The genetics of Primula sinensis.

IV. Indications as to the autogenetic relationship of leaf and inflo-
rescence. Ann. Bot. 49.

and W. H. Eyster. (1928). Pericarp studies in maize. III. The

frequency of mutation in variegated maize pericarp. Genetics 13,

111-120.
Andersson-Kottoe, I. (1930). Variegation in three species of ferns

(Polystichum angulare, Lastraea atrata, and Scolopendrium vulgare).

Zeitschr. ind. Abstl. 56.

. (1931). The genetics of ferns. Bibliographia Genetica 8, 269-294.

Artom, C. (1920). Nuovi fatti e nuovi problemi sulla biologia e sulla

sistematica del genere Artemia. R. Accad. Lincei 29.
. (1926). Tetraploidismo e Gigantismo. Esame comparativo

degli stadi postembrionali dell' Artemia salina diploide e tetraploide.

Int. Rev. d. ges. Hydrobiol. u. Hydrographie 16.
Astatjroff, B. L. (1929). Studien liber die erbliche Veranderung der

Halteren bei Drosophila melanogaster Schin. Arch. Entwicklgmech.

115, 424-447.
. (1930). Analyse der erblichen Storungsfalle der bilateralen

Symmetrie im Zusammenhang mit der selbstandigen Variabilitat

ahnlicher Strukturen. Zeitschr. ind. Abstl. 55, 183-262.
Auerbach, Charlotte. (1935-1936). The development of the legs,

wings, and halteres in wild type and some mutant strains of Drosophila

melanogaster. Trans. Roy. Soc. Edinburgh 58, Pt. III.

B

Bagg, H. J. (1926a). Hereditary abnormality of the viscera I. Amer.

J. Anat. 36.
. (19266). Hereditary abnormalities of the limbs, their origin

and transmission II. Amer. J. Anat. 43.

317

.318 PHYSIOLOGICAL GENETICS

Balkaschina, E. I. (1929 . Bin Fall der Erbhomoosis (die Genovariation

"AriBtopedia") 1 >»-i Drosophila melanogaster. Arch. Enturicklgmech.

115, I In 463.
Baltzbr, F. (1933). Uber die Entwicklung von Triton-Bastarden ohne

Eikern. Verhandl. Deutsch. Zool. Oea.

inn! V. Db Roche. I 1936). uber die Entwicklungsfaehigkeil hap-

toider Triton alpestris-Keime und ueber die Aufhebung der Entwick-

lungshemmung bei Geweben letaler bastardmerogonischer ECombina-

tionen durch Transplantation in einen normalen Wirt. Rev. Suisse

Zool. 43.
Baron, A. L. (1935). Facel number in Drosophila melanogaster as

influenced by certain genetic and environmental factors. Jour. Exp.

Zool. 70.
BARROWS, E. F. (1934). Modification of the dominance of agouti to

non-agouti in tin" mouse, .lour, Gem tics 29.
I '• \ i u, E. (1924). Untersuchungen iiber das Wesen, die Entstehung und

die Vererbung von Rassenuntersehieden bei Antirrhinum tnajus.

Kihlioth. (laid. 4.

— . (1926a). Untersuchungen iiber Faktorinutationen. I. Antir-

rhinum ma jus mut. phantastica. Zeitschr. ind. Abstl. 41.

— . (19266). Untersuchungen iiber Faktorinutationen. II— III.

Zeitschr. ind. Abstl. 41.

(1932). Artumgrenzung und Artbildung in der Gattung Antir-

rhinum, Section Antirrhinastrum. Zeitschr. ind. Abstl. 63.

Beadle, G. W. (1930). Genetical and cytological studies of mendelian
asynapsis in Zea mays. Cornell Univ. Agric. Exper. Sta. Mem. 129.

. (1931). A gene in maize for supernumerary cell divisions follow-
ing meiosis. Cornell Univ. Agric. Exper. Sta. Mem. 135.

. (1932a). A gene for sticky chromosomes in Zea mays. Zeitschr.

ind. Abstl. 63, 195-217.

. (19326). A gene in Zea mays for failure of cytokinesis during

meiosis. Cytologic! 3, 142-155.

and Boris Ephrussi. (1936). The differentiation of eye pigments
in Drosophila as studied by transplantation. Genetics 21.

and . (1937). Development of eye colors in Drosophila:

Diffusible substances and their interrelations. Genetics 22.

Becker, G. (1931). Experimentelle Analyse der Genom- und Plasmon-
wirkung bei Moosen. III. Osmotischer Wert heteroploider Pflanzen.
Zeitschr. ind. Abstl. 60, 17-38.

BECKER, Erich. (1937). Die Rotbrauue Zeichnung der Wespennest-
niuetter, cine durch mechanischen Beiz ausgeloeste Pigmentablagerung
in Liesegangschen Ringen. Zeitschr. f. Vagi. Physiol. 24.

Beebe, C. W. (1907). Geographic variations in birds, with special refer-
ence to the effects of humidity. Zoologica. N. Y. Zool. Soc.

liEi.i.\.iEFK, N. K. (1931). ErbUche Asymmetrie bei Drosophila. Ein
neues Gen im IV. Chromosom von Drosophila melanogaster. Biol.
Zeniralbl. 51, 701-708.

Berg, R. (1934). The relative mutation frequencies in Drosophila
chromosomes. Comp. Rend. Acad. Sci. URSS. 4.

BIBLIOGRAPHY 319

Bergmann, M. (1936). Proteins and Proteolytic Enzymes. Harvey
Lectures 1935-1936.

and C. Niemann. (1937). Newer Biological aspects of Protein

Chemistry. Science 86.

Berndt, W. (1925). Vererbungsstudien an Goldfischrassen. Zeitschr.
ind. Abstl. 36.

Blakeslee, A. F. (1921). Types of mutation and their possible signifi-
cance in evolution. Amer. Naturalist 55.

. (1922). Variations in Datura due to changes in chromosome

number. Amer. Naturalist 56.

Blunn, C. T., and P. W. Gregory. (1935). The embryological basis of
size inheritance in the chicken. Jour. Exp. Zool. 70, 397-414.

Bobrow, A. S. (1930). Experimentelle Untersuchungen am Spinner
Orgyia antiqua L. Inst. Exp. Biol. Moskau 10.

. (1937). The development of eye colors in Drosophila as studied

by transplantation. Amer. Naturalist 71.

Bodenstein, D. (1933). Beintransplantationen an Lepidopteren Raupen.
I. Transplantationen zur Analyse der Raupen- und Puppenhaeutung.
Arch. Entwicklgmech. 128.

. (1936). Das Determinationsgeschehen bei Insekten mit Aus-

schluss der fruehembryonalen Determination. Ergebnisse d. Biol. 13.

Bonnevie, Kristine. (1930-1931). Vererbbarer Cerebrospinaldefekt (?)
bei Mausen, mit sekundaren Augen- und Fuss-Anomalien, nebst
Turmschadelanlage. Vorlaufige Mitteilung (with an English sum-
mary). Avh. Norske Vid. Akad. Oslo I. Mat.-Nat. Kl. 13.

. (1932). Epidermispolster der menschlichen Finger und Zehen.

Die vererbbaren Kopf- und Fuss-anomalien der Little und Baggschen
Mauserasse in ihrer embryologischen Bedingtheit. Zeitschr. ind.
Abstl. 62, 72-84.

. (1934). Embryological analysis of gene manifestation in Little

and Bagg's abnormal mouse tribe. Jour. Exp. Zool. 67, 443-520.

— . (1935). Vererbbare Missbildungen und Bewegungstoerungen

auf embryonale Gehirnanomalien zurueckfuehrbar. Erbarzt. Nr. 10.

. (1936a). Abortive differentiation of the ear vesicles following a

hereditary brain anomaly in the short-tailed waltzing mice. Genetica
18.

(19366). Pseudencephalie als Spontane Recessive (?) Mutation

bei der Hausmaus. Vid.-Akad. Skr. I. Mat.-Naturv. Kl. 9.

— . (1936c). Vererbbare Gehirnanomalie bei kurzschwaenzigen Tanz-

maeusen. Act. Path. Micr. Scand. Suppl. 26.
Bonnier, G. (1928). A Drosophila mosaic, probably due to dispermic

fertilisation. Jour. Genetics 19, 257-260.
Boveri, Th. (1904). Protoplasmadifferenzierung als ausloesender Faktor

fuer Kernverschiedenheit. Sitzber. Phys. Med. Ges. Wuerzburg.

— . (1918). Zwei Fehlerquellen bei Merogonieversuchen. Arch.

Entwicklgmech. 44.
Boycott, A. E., C. Diver, S. L. Garstang, and F. M. Turner. (1930).

The inheritance of sinistrality in Limnaea peregra (Mollusca, Pul-

monata). Phil. Trans. Roy. Soc. London B 219.

320 PHYSIOLOGICAL GENETICS

Braun, Wkknkk. (1936). I'I'tT das Zellteilungamuater im Puppen-
fiuegelepithel der Mehlmotte Ephestia kuehniella Z. in seiner Beziehung
zur Auabildung des Zeichnungamusters. Arch. Entwicklgmech. 136.
Brjeitbnbecher, .F. K. (1025). The inheritance <>f sex-limited bilateral

asymmetry in Bruchua. Genetics 10.
Bridges, C. B. (1913). Non-diajunction of the sex-chromosomes of
Droaophila. Jour. Exp. Zool. 16.

— . (1010a). The genetics of purple eye color in Droaophila melano-
H.-ister. Jour. Exp. Zool. 28.

— . (191%). Specific modifiers of cosin eye color in Droaophila
melanogaater. Jour. Exp. Zool. 28.
— . (1919c). Vermilion-deficiency. Jour. Gen. Physiol. 1.

. (1921a). Genetical and cytological proof of non-disjunction of

the fourth chromosome of Drosophila melanogaater. Proc. Nat. Acad.
Sci. 7.
. (19216). Triploid intersexes in Drosophila melanogaster. Sci-
ence, N. S. 54.

. (1922). The origin of variations in sexual and sex-limited

characters. Amer. Naturalist 66.

. (1925a). Sex in relation to chromosomes and genes. Amer.

Naturalist 69.
— -. (19256). Elimination of chromosomes due to a mutant (mi-

nute-n) in Drosophila melanogaster. Proc. Nat. Acad. Sci. 11.
. (1928). The behavior of mutable genes. A review of the paper

by Dr. M. Demerec. Coll. Net. 3.
— . (1930). Haploid Drosophila and the theory of genie balance.

Science 72, 405-406.
— . (1936). The Bar "gene" a duplication. Science 83.

— and Th. Dobzhansky. (1933). The mutant "proboscipedia" in
Drosophila melanogaster — a case of hereditary homoosis. Arch.
Entwicklgmech. 127, 575-590.

— and T. H. Morgan. (1923). The third-chromosome group of

mutant characters of Drosophila melanogaster. Carnegie Inst. Wash-
ington, Publ. 327.

Brieger, F. G., and A. G. Mangelsdorf. (1926). Linkage between a
flower color factor and self-sterility factors. Proc. Nat. Acad.
Sci. 12.

Brink, R. A. (1929a). An enzyme difference associated with the waxy
gene in maize. Genetics 14, 569-590.

. (19296). Studies on the physiology of a gene. Quart. Rev. Bio.

4, 520-543.

. (1932). Are the chromosomes aggregates of groups of physio-
logically interdependent genes? Amer. Naturalist 66, 444-451.

and C. R. Burnham. (1927). Differential action of the sugary

gene in maize on two alternative classes of male gametophytes. Genetics
12, 348-378.
Buddenbrock, W. v. (1928). Grundriss der vergleichenden Physiologic.
Berlin.

BIBLIOGRAPHY 321

— . (1930). Beit rag zur Histologie und Physiologfe der Raupen-
haoutung mit besonderer Beruceksichtigung der Versonschen Druesen.
Zeitschr. f. Morph. u Oekol. 18.
— . (1931). Untersuehungen ueber die Haeutungshormone der

Schmetterlingsraupen. II. Zeitschr. vergl. Physiol. 14.
Bytinsky-Salz, H. (1933). Untersuehungen an Lepidopteren-Hybriden.
II. Entwicklungsphysiologische Experimente ueber die Wirkung der
disharmonischen Chromosomenkombinationen. Arch. Entwicklgmech.
129.

Caspari, E. (1933). Uber die Wirkung eines pleiotropen Gens bei der

Mehlmotte Ephestia ktihniella Zeller. Arch. Entwicklgmech. 130, 353-

381.
. (1936). Zur Analyse der Matroklinie der Vererbung in der a-Serie

der Augenfarbenmutationen bei der Mehlmotte Ephestia kuehniellaZ.

Zeitschr. ind. Abst. Vererb. 71.

and E. Plagge. (1935). Versuche zur Physiologie der Verpuppung

von Schmetterlingsraupen. Naturiviss.

Castle, W. E. (1919). Is the arrangement of the genes in the chromosome
linear? Proc. Nat. Acad. Sci. 5.

. (1932). Growth rates and racial size in rabbits and birds. Sci-
ence 76, 259-260.

and P. W. Gregory. (1929). The embryological basis of size

inheritance in the rabbit. Jour. Morph. Physiol. 48, 81-103.

and P. W. Gregory. (1931). The effects of breed on growth of the

embryo in fowls and rabbits. Science 73, 680-681.

Chen, T. (1929). On the development of imaginal buds in normal and
mutant Drosophila melanogaster. Jour. Morphol. Physiol. 47, 136-
187.

Chesley, P. (1935). Development of the short-tailed mutant in the house
mouse. Jour. Exp. Zool. 70, 429-459.

Child, Chr. J. (1925). Senescence and rejuvenescence. Chicago.

Child, G. (1935). Phenogenetic studies on scute-1 of Drosophila melano-
gaster. I. The associations between the bristles and the effects of
genetic modifiers and temperature. II. The temperature-effect period.
Genetics 20, 109-126, 127-155.

. (1936). Phenogenetic studies in Scute of D. melanogaster. III.

The effect of temperature on scute 5. Genetics 21.

Chittenden, R., and P. Pellew. (1927). A suggested interpretation
of certain cases of Anisogeny. Nature 119.

Clausen, R. E., and T. H. Goodspeed. (1925). Interspecific hybridiza-
tion in Nicotiana II. A tetraploid glutinosa-tabacum hybrid; an experi-
mental verification of Winge's hypothesis. Genetics 10.

and W. E. Lammerts. (1929). Interspecific hybridization in

Nicotiana. X. Haploid and diploid merogony. Amer. Naturalist 63.

Cochrane, Flora. (1936). Observations on eye-colour development
in Drosophila pseudo-obscura. Jour. Genetics 32.

322 PHYSIOLOGICAL GENETICS

Co< 8 \i m:, E. A. (1935). The origin <>l' gynandromorpha in the Lepidop-

tera from binucleate ova. Trans. Roy. I^niom. Soc, London 83.
Conklin, E.G. (1903). The cause of inverse symmetry. A nat. An:. 23.
. (1925). Cellular differentiation in: Cowdry's General Cytology.

[Jnivereity of Chicago Press.

(1929 . Problems of development. (7th Sedgwick Memorial

Lecture Woods Hole 1928.) Amer. Naturalist 63, 5-36.
Correns, C. (1899). Untersuchungen iiber die Xenien bei Zea Mays.

Ber. deutsch. Hot. Ges. 17.
— . (1905). Zur Kenntnis der scheinbar neuen Merkmale der

Bast aide. Ber. deutsch. Bot. Ges. 23.
. (1910). Der Uebergang aus dem homozygotischen in eiueu

hetetnzvgot ischeu Zustand im selben Individuum bei luiiithlaettrigen

und gestreiftbluehenden Mirabilis-Sippen. Ber. deutsch. Bot. Ges. 28.

— . (1919). Vererbungsversuche mit buntblaettrigen Sippen. I.

Capsella Bursa pastoris albovariabilis und chlorina. Sitzb. Preuss.

Akad. Wiss. 34.

— . (1928). Uber nicht mendclnde Vererbung. Zeitschr. ind. Abstl.

Suppl.-Bd. 1.

Crampton, H. E. (1924). The coincident production of dextral and
sinistral young in the land-gasteropod Partula. Science 59.

Ckew, F. A. E., and Rowena Lamy. (1932). A case of conditioned
dominance in Drosophila obscura. Jour. Genetics 26, 351-358.

Csik, L. (1934a). Die Zusammenarbeit einiger Gene bei der Determina-
tion der Fliigelgrosse von Drosophila melanogaster. Biol. Zentralbl.
64, 614-645.

. (19346). Die Zusammenwirkung einiger Borsten-Gene auf die

Lange der Borsten bei Drosophila melanogaster. (Vorl. Mittlg.)
(Hungarian, German summary.) Arb. II. Abtlg. Ungar. Biol. Forsch.
Inst. 7, 325-329.

Cuenot, L. (1902). La loi de Mendel et l'heredite de la pigmentation
chez les souris. Arch. Zool. Exp. et Gen. Notes et Revue.

. (1911). Les determinants de la couleur chez les souris, 6tude

comparative. Arch. Zool. Exp. 8.

Cunningham, I. T. (1908). Hormonentheorie der Vererbung basiert auf
den Tatsachen ueber den Einfluss der Gonadenhormone auf sekundaere
Geschlechtscharaktere. Arch. Entwicklungmech. 26.

D

Dahlberg, G. (1929). Genotypische Asymmetrien. Zeitschr. ind. Abstl.

53, 133-148.
Danforth, C. H. (1929). The effect of foreign skin on feather pattern

in the common fowl (Callus domesticus). Arch. Entwicklgmech. 116,

242-252.
. (1932). Artificial and hereditary suppression of sacral vertebrae

in the fowl. Proc. Soc. Exp. Biol. Med. 30.

— . (1933). Genetic factors in the response of feather follicles to

thyroxin and theelin. Jour. Exp. Zool. 66, 183-196.

BIBLIOGRAPHY 323

Danneel, R. (1934). Zur Physiologie dcr Kaelteschwaerzung beim

Russenkaninchen I. Biol. Zenlralbl. 54.
. (1936). Die Facrbung unserer Kaninchenrassen und ilire histo-

genetischen Grundlagen. Zeitschr. ind. Abstl. 71.

and E. Lubnow. (1936). Zur Physiologie der Kaelteschwaerzung

beim Russenkaninchen. Biol. Zenlralbl. 66.
Darby, Hugh H. (1934). The mechanism of asymmetry in the Alpheidae.

Carnegie Inst. Washington Publ. 435.
■ . (1935). The mechanism of chela differentiation in the Crustacea.

Carnegie Inst. Washington Publ. 452.
Davenport, C. B. (1908). Determination of dominance in Mendelian

inheritance. Proc. Amer. Phil. Soc.
. (1935). A provisional hypothesis of child development. Proc.

Amer. Phil. Soc. 75.
David, Lorna T. (1932). The external expression and comparative

dermal histology of hereditary hairlessness in mammals. Zeitschr.

Zellforsch. 14, 616-719.
Dawson, A. B. (1935). The influence of hereditary dwarfism on the

differentiation of the skeleton of the mouse. Anat. Rec. 61.
Dellingshausen, M. v. (1935). Entwicklungsgeschichtlich-genetische

Untersuchungen an Epilobium. V. Permeabilitaetsstudien an zwei

genetisch verschiedenen Plasmen. Planta 23.
De Meijere, J. C. H. (1910). Ueber Jacobsens Zuechtungsversuche usw.

Zeitschr. ind. Abstl. 3.
Demerec, M. (1928a). The behavior of mutable genes. Zeitschr. ind.

Abstl. 1928 Suppl. Bd. 1, 183-193.
. (19286). Mutable characters of Drosophila virilis. I. Reddish-
alpha body character. Genetics 13, 359-388.
. (1929a). Changes in the rate of mutability of the mutable

miniature gene of Drosophila virilis. Proc. Nat. Acad. Sci. 15, 870-876.
. (19296). Genetic factors stimulating mutability of the miniature-
gamma wing character of Drosophila virilis. Proc. Nat. Acad. Sci. 15,

834-838.
. (1929c). Mutable genes in Drosophila virilis. Proc. Internal.

Congr. Plant Sciences 1, 943-946.
. (1930a). The behaviour of two mutable genes of Delphinium

ajacis. Proc. 5th Internal. Bot. Congr. Cambridge 196-197.
. (19306). A genetic factor affecting germinal mutability of minia-
ture-alpha wing character of Drosophila virilis. "The laws of life."

Mem. Vol. 60th birthday Prof. Ruzicka, 45-56.
. (1931a). Behaviour of two mutable genes of Delphinium ajacis.

Jour. Genetics 24, 179-193.

— . (19316). The gene. Abstract: The Biol. Labor. 3, 29-32.
. (1932a). Effect of temperature on the rate of change of the

unstable miniature-3 gamma gene of Drosophila virilis. Proc. Nat.
Acad. Sci. 18, 430-434.

— . (19326). Rate of instability of miniature-3 gamma gene of
Drosophila virilis in the males in the homozygous and in the hetero-
zygous females. Proc. Nat. Acad. Sci. 18, 656-658.

324 PHYSIOLOGICAL GENETICS

. (1933). What is a gene? Jour. Heredity 24, 369-378.

. (1934a). Biological action of small deficiences of X-chromosome

of Drosophila melanogaster. I'roc. Nat. Acad. Sci. 20, 354-359.
— . (19346). Effecl of X-raya on the rate of change in the unstable
miniature-3 gene of Drosophila virilis. Proc. Nat. Acad. Sci. 20, 28-31.
— . (1934c). The gene and its role in ontogeny. Cold Spring Harbor

Symposia <>n Quant. Biol. 2, 110-115.

— . (1935a). Role of genes in evolution. Arner. Naturalist 69.

— . (1935b). Unstable genes. But. Rev. 1.

de Viues, H. (1910). The mutation theory. Vol. 2 (German original

published in 1903.)
Dobrovolskaia-Zavadskaia, N. (1928). L'irradiation tics testicules el

l'heredite chez la souris. Arch. Biol. 38.

— . (1929). The problem of species in view of the origin of some new-
forms in mice. Biol. Revs. 4.

and N. Kobozieff. Sur le facteur lethal accompagnant I'anourie el

la brachyurie chez la souris. Compt. Rend. Acad. Sci., Paris 191.
Dobzhansky, Th. (1929). The influence of the quantity and quality

of chromosomal material on the size of the cells in Drosophila melano-
gaster. Arch. Entwicklgmech. 115.
. (1930). The manifold effect of the genes Stubble and stubbloid in

Drosophila melanogaster. Z. ind. Abstl. 54.
. (1931a). Interaction between female and male parts in gynan-

dromorphs of Drosophila simulans. Arch. Entwicklgmech. 123.
. (19316). Studies on the intersexes and supersexes in Drosophila

melanogaster. (Russian, English summary.) Bull. Bur. Genet. 8.
. (1932). The baroid mutation in Drosophila melanogaster.

Genetics 17.
. (1935). Maternal effect as a cause of the difference between the

reciprocal crosses in Dros. pseudoobscura. Proc. Nat. Acad. Sci. 21.

. (1936). Position effects on genes. Biol. Rev. 11.

and F. N. Duncan. (1933). Genes that affect early developmental

stages of Drosophila melanogaster. Arch. Entwicklgmech. 130, 109-128.
and A. H. Stubtevant. (1931). Translocations between the

second and third chromosomes of Drosophila and their bearing on

Oenothera problems. Carnegie Inst. Washington Publ. 421.

and . (1932). Change in dominance of genes lying in

duplicating fragments of chromosomes. Proc. 6th Internal. Congr.
Genetics 2.

and . (1935). Further data on maternal effects in Dros.

pseudoobscura hybrids. Proc. Nat. Acad. Sci. 21.
Dobbies-Rugeb, Kate. (1929). Experimentelle Analyse der Genom-

und Plasmonwirkung bei Moosen. I. Teilungsgeschwindigkeit.

Zeitschr. ind. Abstl. 52, 390-405.
Donald, H. P. (1936). On the suppression of Tangled in Drosophila

pseudo-obscura. Proc. Roy. Soc. Edinburgh 56.
Dobfmeisteb, G. (1863-1864). Uber Arten und Varietaeten der Schmet-

terlinge. Mitt. Naturwiss. Verein. Steiermark.

BIBLIOGRAPHY 325

— . (1864). t)ber die Einwirkung verschiedener waehrend der
Entwicklungsperioden angewendeter Waermegrade auf die Faerbung
und Zeichnung der Schmetterlinge. Mitt. Naturwiss. Verein Steiermark.
(1879). Uber den Einfluss der Temperatui bei der Erzeugung der

Schmetterlings-varietaeten. Mitt. Naturwiss. Verein. Steiermark.

Driver, E. C. (1926). The temperature effective period, the key to eye-
facet number in Drosophila. Jour. Exp. Zool. 46.

. (1931 a). Temperature-effective periods in reciprocal crosses in

Drosophila melanogaster. Jour. Exp. Zool. 59.

. (1931 b). Temperature and gene expression in Drosophila.

Jour. Exp. Zool. 59.

Dry, F. W. (1928). The agouti coloration of the mouse (Mus museums)
and the rat (Mus norvegicus). Jour. Genetics 20.

Dubinin, N. P. (1929a). Allelomorphentreppen bei Drosophila melano-
gaster. Biol. Zentralbl. 49, 328-339.

. (19296). Studies on the phenomenon of step allelomorphism 1.

Allelomorphs scute1, scute2, scute3. Jour. Exp. Biol. 5.

. (1932a). Step-allelomorphism and the theory of centres of the

gene, achaete-scute. Jour. Genetics 26, 37-58.

. (19326). Step-allelomorphism in Drosophila melanogaster. The

allelomorphs achaete-scute1 °, achaete-scute11, and achaete3-scute13.
Jour. Genetics 25, 163-181.

■ . (1935). Discontinuity and continuity in the structure of the

hereditary materials. Trud. Dinam. Razvit. 10.

. (1936). A new type of position effect. Inst. Biol. Exp. USSR 5.

and H. Friesen. (1932). Die Unmoglichkeit einer Erklarung des

Falls der Treppenallele scute vom Standpunkte der Goldschmidtschen
physiologischen Theorie der Vererbung. Biol. Zentralbl. 52, 147-162.

and S. J. Goldat. (1936). A new case of Bar mutation. Inst.

Biol. Exp. USSR 5.

and M. A. Heptner. (1935). A new phenotypic effect of the

Y-chromosome in Drosophila melanogaster. Jour. Genetics 30.

and B. N. Sidorov. (1934). Relation between the effect of a gene

and its position in the system. Amer. Naturalist 68.
Du Bois, A. M. (1929). Recherches morphologiques sur le developpement

de l'appareil copulateur des femelles intersexuees de Lymantria dispar

L. Arch. Entwicklgmech. 124.
Duncker, G. (1904). Symmetrie und Asymmetrie bei bilateralen

Tieren. Arch. Entwicklgmech.
Dunn, L. C. (1925). The inheritance of rumplessness in the domestic

fowl. Jour. Heredity 16.
. (1934). Analysis of a case of mosaicism in the house-mouse.

Jour. Genetics 29.
. (1937). Studies on spotting patterns II. Genetic analysis of

variegated spotting in house mice. Genetics 22.
and D. R. Charles. (1937). Studies on spotting patterns I.

Analysis of quantitative variations in the pied spotting of the house

mouse. Genetics 22.

326 PHYSIOLOGICAL GENETICS

:u id Jeanne Gptne. I L935 1. The relationship between the effects
oi certain mutations on developmental rate and on adull characters.
Biol. Zentralbl. 66, 385 389.

and Coyne, M. J. (1937 I. The effects of i he minute mutations of
Drosophila melanogaster on developmental rate. Hereditas 28.

.•mil W I.\m>\i br. (1934). The genetics of the rumpless fowl.

.1 mil . In in ties 29.

.■mil — . (1930). Further data on genetic modification of

rumplessness in the fowl. Jour. Genetics 33.

— , E. C. MacDowell, and G. A. Lebedeff. (1937). Studies on

spotting patterns. 111. Interaction between genes affecting white

spotting and those affecting color in the house mouse, (lenities 22.

Du Toit, P. I. (1913). Untersuchungen iiber das Synsacrum und den
Schwanz von Gallus domesticus nebst Beobachtungen ueber Schwanz-
losigkeit l>ei Kaulhuehnern. Jena Zeitscht. Natunr. 49.

E

East, E. M. (1929). The concept of the gene. I'roc. Internal. Congr.

Plant Sci. 1.

— . (1934). The nucleus-plasma problem. A mer. Naturalist 63.
Eigenbrodt, Harold J. (1930). The somatic effects of temperature on

a homozygous race of Drosophila. Physiol. Zool. 3.
Eker, R. (1935). The shortwing gene in Drosophila melanogaster and the

effect of temperature on its manifestation. Jour. Genetics 30.
Emerson, R. A. (1914). The inheritance of a recurring variation in

variegated ears of maize. Amer. Naturalist 48.
. (1917). Genetical studies on variegated pericarp in maize.

Genetics 2.
. (1929). The frequency of somatic mutation in variegated

pericarp of maize. Genetics 14, 488-5 1 1 .
Emsweller, S. L., and H. A. Jones. (1935). A gene for control of inter-
stitial localization of chiasmata in Allium fistulosum L. Science 81.
Engelsmeier, W. (1935). Nachweis der alternativen Modifikabilitat der

Haarfiirbung beim Russenkaninchen. Ein Beitrag zur Theorie der

quantitative!! Wirkung der Gene. Zeilschr. ind. Abstl. 68, 301-410.
Ephrussi, B. (1934). The absence of autonomy in the development of the

effect of certain deficiencies in Drosophila melanogaster. Proc. Nat.

Acad. Sci. 20.
. (1935). The behavior in vitro of tissues from lethal embryos.

Jour. Exp. Zool. 70.
and G. W. Beadle. (1937a). DeVeloppement des couleurs des

yeux chez la Drosophile: Influence des implants sur la couleur des yeux

de I'hote. Bull. Biol. 71.
and . (19376). Dcveloppement des couleurs des yeux, chez

la Drosophila: Revue des experiences de transplantation. Bull. Biol.

71.
— and . (1937c). Development of eye colors in Drosophila:

Transplantation experiments on the interaction of vermilion with other

eye color mutants. Genetics 22.

BIBLIOGRAI'H V 327

— and . (1937d). Development of eye colors in Drosophila:

Studies of the mutant claret. Jour. Genetics 31.
Euler, H. v. (1929). Chemische Untersuchungen an Chlorophyll-
mutanten, I. Hereditas 13, 61-79.

, B. Bergman, H. Hellstrom, and D. Bcrstrom. (1935). Kon-

stanz des Chlorophyllgehaltes und Chromatophorendegeneration chloro-
phyllmutierender Gerstensippen. Hereditas 21, 119-128.

, Dagmar Btjrstrom, and H. Hellstrom. (1933). Uber die

Konstanz des Chlorophyllgehaltes in drei Chlorophyllmutanten.
Hereditas 18, 225-244.

— , A. Forssberg, D. Runehjelm, and H. Hellstrom. (1931).
Zur chemischen Charakterisierung von erblichen Chlorophylldefekten.
Zeitschr. ind. Abstl. 59, 131-152.

and H. Hellstrom. (1933). Uberein Indolderivat aus zwei chloro-

phyllmutierenden Gerstensippen. Hoppe-Seyler's Zeitschr, physiol.
Chemie 217, 23-27.
and H. Nilsson. (1929). Enzymchemische Vererbungsstudien.

I. Arkiv Kemi, Mineralogi, Geologi 10B, No. 6.
Eyster, W. H. (1924). A genetic analysis of variegation. Genetics 9.
. (1928). The mechanism of variegations. Zeitschr. ind. Abstl.

Suppl. Bd. 1, 666-686.
. (1929). The bearing of variegations on the nature of the gene.

Proc. Internal. Congr. Plant Sci. 1.

F

Federley, H. (1907). Lepidopterologische Temperaturexperimente.

Festschr. Palmen, Helsingfors.
. (1913). Das Verhalten der Chromosomon bei der Spermatogenese

der Schmetterlinge usw. Zeitschr. f. ind. Abstl. 9.

(1920). Die Bedeutung der polymeren Faktoren fuer die Zeich-

nung der Lepidopteren. Hereditas 1.
Feldotto, W. (1933). Sensible Perioden des Fluegelmusters bei Ephestia

kuehniella Zeller. Arch. Entwicklgmech. 128.
Feldstein, M. J. and A. H. Hersh. (1935a). The calculation of relative

growth constants. Amer. Naturalist 69.
and . (19356). The determination of genetic constants of

relative growth. Amer. Naturalist 69.
Ferwerda, F. P. (1928). Genetische Studien am Mehlkiifer Tenebrio

molitor L. Genetica 11, 1-110.
Fischer, E. (1895). Transmutation der Schmetterlinge infolge der

Temperaturaenderungen. Experiment elle Untersuchungen liber die

Phylogenese der Vanessen. Zurich.
Fischer, Eu. (1933). Genetik und Stammesgeschichte der menschlichen

Wirbelsaule. Biol. Zentralbl. 53, 203-220.
Fisher, R. A. (1928). The possible modification of the response of the

wild type to recurrent mutations. Amer. Naturalist 62.
. (1930). The genetical theory of natural selection. London.

328 PHYSIOLOGICAL GENETICS

. (1931). The evolution of dominance. Biol. Revs. 6, 345-368.

. (1932). Tlir evolutionary dification of genetic phenomena.

Proc. (till I nli runt. Conor. Genetics 1, 105-172.

— . (1935). Dominance in poultry. Phil. Trans. Roy. Soc. London B

226.
Fribsbn, II. (1936). Roentgenmorphosen bei Drosophila. Arch. Ent-

wicklgmechan. 134.
Fryer, J. C. .1. (1913). An investigation by pedigree breeding into the

polymorphism of Papilio polytes Linn. Phil. Trans. Roy. Soc. 204.
I i mi, K. (1927). On the color pattern produced by various agents in the

goldfish. Folia Anat. Japon. 5.

— . (1930). The definite localization of the color pattern in the

goldfish. Folia Anat. Japon. 8.
FORD, E. B. (1930). The genetics of Papilio dardanus Brown (Lep.).

Trans. Roy. Entomol. Soc. London 85.

— and J. S. Huxley. (1927). Mendelian genes and rates of develop-
ment in Gammarus chevreuxi. Brit. Jour. Exper. Biol. 6, 112-134.

- and . (1929). Genetic rate-factors in Gammarus. Arch.

Entwicklgmech. 117, 07-79.
Fraenkel, B. (1934). Pupation in flies initiated by a hormone. Nature

133.
. (1935). A hormone causing pupation in the blowfly Calliphora

Erythrocephala. Proc. Royal Soc. London Ser. B. 118.
Fraps, Richard M., and Mary Juhn. (1930a). Developmental analysis

in plumage. II. Plumage configurations and the mechanism of

feather development. Phys. Zool. 9.
and . (193Gb). Developmental analysis in plumage. III.

Field functions in the breast tracts. Phys. Zool. 9.
Friesen, H. (1932). Kettenmutationen. Zeitschr. Ind. Abstl. 62,

504-534.

G

Gabritschewsky, E. and C. B. Bridges. (1928). The giant mutation in

Drosophila melanogaster. Zeitschr. Ind. Abstl. 46.
Gabritschewsky, E. (1930). Die Wirkung der Temperatur auf die

Struktur und die Manifestation des Gens "Riese" (Giant) bei ange-

brannten Larven von Drosophila melanogaster. Arch. Entwicklgmech.

121, 466-409.
Gairdner, A. E. (1929). Male sterility in flax II. Jour. Genetics 21.
Gates, R. R. (1933). The general nature of the gene concept. Nature

132, 708.
Geigy, R. (1926). Une anomalie non-hercditaire provoquee par les rayons

ultra-violets chez la Drosophile. Com/it. Rend. Soc. Phys., Hist. Nat.

Geneve 43, 143-145.
. (1927). Production d'anomalies par exposition de drosophiles

mux rayons ultraviolets. Actes Soc. Helv. Sci. Nat. Bdle IP partie,

191-192.

BIBLIOGRAPHY 329

Gebhardt, F. A. (1912). Die Hauptzuege der Pigmentverteilung im
Schmetterlingsfluegel im Lichte dor Liesegangschen Niederschlaege in

Kolloiden. Verhandl. Dcutsch. Zool. Ges. 22. Jahrcsverslg.
Gerodld, J. H. (1911). The inheritance of polymorphism and sex in

Colias philodice. Arner. Naturalist 45.
. (1923). Inheritance of white wing color, a Bex-limited (sex-
controlled) variation in yellow pierid butterflies. Genetics 8.
Giersberg, H. (1929). Die Fiirbung der Schmetterlinge I. Zeitschr.

vgl. Physiol. 9, 523-552.
Glass, H. B. (1933a). A new allelomorphic compound presenting the

phenotype of the wild Drosophila melanogaster. Jour. Genetics 27,

233-241.
. (19336). A study of dominant mosaic eye-colour mutants in

Drosophila melanogaster. II. Tests involving crossing-over and

non-disjunction. Jour. Genetics 27.
. (1934a). The effect of lethal genes on the non-disjunctional

classes in mutual translocations of Drosophila. Zeitschr. f. ind. Abst.-u.

Vererb. 67.
. (19346). A study of dominant mosaic eye-color mutants in

Drosophila melanogaster. Amer. Naturalist 68.

Godlewski, E. jun. (1909). Das Vererbungsproblem im Lichte der
Entwicklungsmechanik betrachtet. Heft 9 der Vortraege und Auf-
saetze liber Entwicklungsmechanik von Roux.

Goernitz, K. (1923). Versuch einer Klassifikation der haeufigsten
Federfaerbungen. Jour. Ornith. Jahrg. 71.

Goldschmidt, R. (1904). Der Chromidialapparat lebhaft funktionieren-
der Gewebezellen. Zool. Jahrb. (Anat.) 21.

. (1911a). Einfuehrung in die Vererbungswissenschaft. 1. Aufl.

Leipzig. W. Engelmann.

. (19116). tJber die Vererbung der sekundaeren Geschlechts-

charaktere. Muench. Med. Wochschr. 49.

. (1912). Erblichkeitsstudien an Schmetterlingen. I. Zeitschr.

ind. Abstl. 7.

. (1914). Erblichkeitsstudien an Schmetterlingen. II. Zeitschr.

ind. Abstl. 11.

. (1916a). Experimental intersexuality and the sex problem.

Amer. Naturalist 50.

. (19166). Genetic factors and enzyme reaction. Science 43.

. (1916c). A preliminary report on further experiments in inherit-
ance and determination of sex. Proc. Nat. Acad. 20.

. (1917a). Crossing-over ohne Chiasmatypie? Genetics 2.

. (19176). A further contribution to the theory of sex. Jour.

Exp. Zool. 22.

. (1917c). Intersexuality and the endocrine aspect of sex. Endo-
crinology 1.

. (1917a"). A preliminary report on some genetic experiments con-
cerning evolution. Amer. Naturalist 52.

:VM) PHYSIOLOGICAL GENETICS

. (1920a). Mechanismus und Physiologic der Geschlechtsbestim-

mung. Berlin. Gebr. Borntrager.

— . (1920b). Die quantitative!) Grundlagen von Vererbung and
Artbildung. Berlin. I. Springer.

— . (1920c). Untersuchungen liber [ntersexualitaet. Zeitschr. ind.
Abstl. 23.

— . (1920a"). Untersuchungen zur Entwicklungsphysiologie dee Flue-
gelmusters der Schmetterlinge. Arch. Entwicklgmech. 47.
— . (1921a). Analyse der Doppelmissbildungen. Arch. Entwicklg-
mech. 47.

— . (19216). Erblichkeitsstudien an Schmetterlingen. III. Zeit-
schr. ind. Abstl. 25.
. (1921c). Zur Entwicklungsphysiologie der [ntersexualitaet.

Naturwiss. 9.
. (1922). Untersuchungen ueber [ntersexualitaet. II. Zeitschr.

ind. Abstl. 29.

— . (1923a). Untersuchungen ueber Intersexualitaet. III. Zeit-
schr. ind. Abstl. 31.
— . (19236). Einige Materialien zur Theorie der abgestimmten

Reaktionsgeschwindigkeiten. Arch. Mikr. Anat. 98.
— . (1924a). Erblichkeitsstudien an Schmetterlingen. IV. Zeitschr.

ind. Abstl. 34.
— . (19246). Untersuchungen zur Genetik der geographischen

Variation. I. Arch. Entwicklgmech. 101.
— . (1925). Bemerkungen iiber triploide Intersexe. Biol. Zentralbl.

46.
— . (1927c). Physiologische Theorie der Vererbung. Berlin. I.

Springer.
— . (1927a1). Weitere morphologische Untersuchungen zum Inter-

sexualitaetsproblem. Zeitschr. f. Morph. u. Oekol. 8.
— . (1928a). Gen und Aussencharakter. Zeitschr. ind. Abstl. Suppl.

I.

— . (19286). The gene. Quart. Rev. Biol. 3.
— . (1929a). Experimented Mutation und das Problem der soge-

nannten Parallelinduktion. Biol. Zentralbl. 49.
— . (19296). Untersuchungen liber Intersexualitaet. IV. Zeitschr.

ind. Abstl. 49.
— . (1929c). Untersuchungen zur Genetik der geographischen

Variation. II. Arch. Entwicklgmech. 116.
— . (1930). Untersuchungen ueber Intersexualitaet. V. Zeitschr.

ind. Abstl. 66.
— . (1931a). Analysis of intersexuality in the gipsy moth. Quart.

Rev. Biol. 6.
— . (19316). Die entwicklungsphysiologische Erklaerung des Falls

der sogen. Treppenallelomorphe. Biol. Zentralbl. 51.
— . (1931c). Neue Untersuchungen iiber die Umwandlung der

Gonaden bei intersexuellen Lymantria dispar L. Arch. Entwicklgmech.

124.

BIBLIOGRAPHY

331

III.

IV

V

1931d). Die sexuellen Zwischenstufen. Berlin. I. Springer.
1932a). Bemerkungen zur Kritik der quantitativen Natur

multipler Allele. Bull. Lab. Genet. Leningrad 9.

19326). Untersuchungen zur Genetik der geograph.
Arch. Entwicklgmech. 126.

1932c). Untersuchungen zur Genetik der geograph.
Arch. Entwicklgmech. 126.

1932<7). Untersuchungen zur Genetik der geograph.
Arch. Entwicklgmech. 126.
1933a). Untersuehungen zur Genetik der geograph.
Arch. Entwicklgmech. 130.

19336). Untersuchungen zur Genetik der geograph.
Arch. Entwicklgmech. 130.

1933c). Genetics and development. Biol. Bull. 63.
1933rf). Some aspects of evolution. Science 28.
1934a). The influence of the cytoplasm upon gene-controlled
heredity. Amer. Naturalist 68.

19346). Lymantria. Bibliogr. Genet. 11.

;i934c). Untersuchungen liber Intersexualitaet. VI. Zeilschr.
ind. Abstl. 67.

VI

VII.

69.

Zool.

Variation.

Variation.

Variation.

Variation.

Variation.

1935a).
19356).

Gen und Ausseneigenschaft. I. Zeitsch. ind. Abstl. 69.
Gen und Ausseneigenschaft. II. Zeitsch. ind. Abstl.

; 1935c).
;i937a).
41.

(19376).
Nature 140.

1938a).

19386).

Gen und Aussencharacter. III.
Gene and character. IV-VIII.

Biol. Centralb. 56.
Univ. California Publ.

Spontaneous chromatin rearrangements in Drosophila.

The theory of the gene. Scienlif. Monthly. (In press.)
Intersexuality and development. Amer. Nat. (In

press,

(1938c). Spontaneous chromatin rearrangements and the theory
of the gene. Proc. Nat. Ac. (In press.)

— and E. Fischer. (1922). Argynnis paphia-valesina, ein Fall
geschlechtskontrollierter Vererbung. Genetica 4.

and . (1927a). Erblicher Gynandromorphismus bei Schmet-

terlingen. Arch. Entwicklgmech. 109.

— and K. Katsuki. Erblicher Gynandromorphismus und soma-
tische Mosaikbildung bei Bombyx mori. Biol. Zentralbl. 47.

— and . (1928a). Cytologie des erblichen Gynandromor-
phismus von Bombyx mori. Biol. Zentralbl. 48.

— and . (19286). Zweite Mitteilung iiber erbliche Mosaik-
bildung und Gynandromorphismus bei Bombyx mori. Biol. Zentralbl.
48.

— and S. Minami. (1923). liber die Vererbung der sekundaeren
Geschlechtscharaktere. Mendelfestschrift, Bruenn.

— and K. Pariser. (1923). Triploide Intersexe bei Schmetterlingen.
Biol. Zentralbl. 43.

332 PHYSIOLOGICAL GENETICS

Goodrich, II. B. (1927). A study of the development of mendelian
characters in Oryzias latipes. .lour. Exp. Zool. 49, 201-287.
— . (1933). One step iii the development <>f hereditary pigmentation
in the fish Oryzias latipes. Biol. Bull. 65, 249-252.
— . (1935). The development of hereditary color patterns in fish.
Amer. Naturalist 69, 267-277.

and Iha B. Hansen. (1931). The postembryonic development of
Mendelian characters in the goldfish, Carassius auratus. Jour. Exp.
Zool. 59, 337 358.

and K. Nicholas. (1933). Scale transplantation in the goldfish

Carassius auratus. Biol. Bull. 65.

Gordon, M., and A. C. Fraser. (1931). Pattern genes in the platyfish.
Inheritance of naturally occurring color pattern variations in the
Mexican killifish, Platypoecilus niaculatus. Jour. Heredity 22, 169-185.

Gortxer, K. A. (1910). Spieglers "white melanin" as related to domi-
nant and recessive white. Amer. Naturalist 44.

— . (1912). On two different types of melanin. Proc. Soc. Exp.
Biol. Med. 9.
— . (1913). Studies on melanin. V. Jour. Amer. Chem. Soc. 35.

GoTTscHKwsKi, G. (1934). Untersuchungen an Drosophila melanogaster
fiber die Umstimmbarkeit des Phanotypus und Genotypus durch
Temperatureinfliisse. Zeitschr. ind. Abstl. 67, 477-528.

Gowen, J. W. (1928). Mutation, chromosome non-disjunction and the
gene. Science 68.

. (1929). On the nature of gene action. Science 70.

— . (1933a). Constitutional effects of the hairless gene in diploid and
triploid Drosophila. Amer. Naturalist 67, 178-180.
— . (19336). Meiosis as a genetic character in Drosophila melano-
gaster. Jour. Exp. Zool. 65, 83-106.

and E. H. Gay. (1933a). Effect of temperature on ever-
sporting eye color in Drosophila melanogaster. Science 77, 312.

— and . (19336). Gene number, kind and size in Drosophila.

Genetics 18, 1-31.

and . (1933c). Eversporting as a function of the Y-chromo-

some in Drosophila melanogaster. Proc. Nat. Acad. Sci. 19, 122-126.

Graubard, M. A. (1933). Tyrosinase in mutants of Drosophila melano-
gaster. Jour. Genetics 27, 199-218.

Green, C. V. (1936). Shifts in expressivity in the heterozygote of a
dominant lethal gene in the mouse. Jour. Exp. Zool. 73.

( iitKENWooD, A. W., and J. S. S. Blyth. (1935). On the relation between
the site of injection of androsterone and the comb response in the
fowl. Jour. Exp. Phys. Zool. 5.

Gregory, P. W., and W. E. Castle. (1931). Further studies on the
embryological basis of size inheritance in the rabbit. Jour. Exp. Zool.
59, 199-211.

and Harold Goss. (1933). Glutathione concentration and

hereditary body size. II. Glutathione concentration in non-
nursed young of six populations of rabbits differing in genetic constitu-
tion for adult size. Jour. Exp. Zool. 66.

BIBLIOGRAPHY 333

— and . (1934). The hypophyseal growth hormone and gluta-
thione concentration. Does the hormone influence the concentration
concurrently with the stimulation of increase in weight? Jour. Exp.
Zool. 69.

— , and V. S. Asmundson. (1935). Glutathione concentration

of White Leghorn and Barred Plymouth Rock embryos after fourteen
days of incubation. Proc. Soc. Exp. Biol. Med. 32.

— , , and . (1936). Glutathione concentration and

hereditary size. V. Comparative studies with Barred Plymouth Rock
and White Leghorn embryos. Jour. Exp. Zool. 73.

Gross, F. (1932). Untersuchungen ueber die Polyploidie und die Vari-
abilitaet bei Artemia salina. Naturwiss. 20.

Grueneberg, Hans. (1936). A case of complete reversion of a chromo-
somal rearrangement in Drosophila melanogaster. Nature 138.

Guthrie, J. (1925). The asymmetry of the small eye condition in eyeless.
Jour. Exp. Zool. 42.

Guyer, M. F. (1911). Nucleus and cytoplasm in heredity. Amer.
Naturalist 45.

H

Haase-Bessell, Gertratjd. (1928). tJber Genombindungen. Verh. 5.

Int. Kongr. Vererb. Bin. 1927. Zeitschr. ind. Abstl. Suppl. Bd. 1,

778-784.
Haberlandt, G. (1935a). Beitraege zum Crataegomespilus-Problem.

Sb. Preuss. Akad. Wiss. Phys.-Math. Kl. 28.
■. (19356). Uber den Blattbau sexueller Bastarde zwischen Mispel

und Weissdorn. Sb. Preuss. Akad. Wiss. Phys.-Math. Kl. 6.
Hadjidimitroff, P. (1933). Die Pigmentverteilung im Kaninchenhaar.

Zeitschr. Zuecht. B, 27.
Hadorn, E. (1934). Die Entwicklungsleistungen bastardmerogonischer

Keimteile von Triton in vitro. Revue Suisse Zool. 41.
. (1936). tjbertragung von Artmerkmalen durch das entkernte

Eiplasma beim merogonischen Triton-Bastard, palmatus-Plasma X

cristatus-Kern. Verhandl. d. Deutsch. Zool. Ges.
Haecker, V. (1890). tJber die Farben der Vogelfeder. Arch. Mikr.

Anal. 35.
— . (1918). Entwicklungsgeschichtlichc Eigenschaf 1 8 analyse

(Phaenogenetik). Jena.
. (1925). Aufgaben und Ergebnisse der Phaenogenetik. Biblioth.

Gen. 1.
and G. Meyer. (1902). Die blaue Farbe der Vogelfedern. Zool.

Jahrb. (Syst.). 15.
Haemmerling, J. (1929). Dauermodifikationen. Handb. Vererb. 11.
. (1934a). I. tlber formbildende Substanzen bei Acetabularia

mediterranea, ihre raeumliche und zeitliche Verteilung und ihre Her-

kunft. Arch. Enticicklgmech. 131.
. (19346). II. Regenerationsversuche an kernhaltigen und kern-

losen Zellteilen von Acetabularia Wettsteinii. Biol. Zentralbl. 54.

334 PHYSIOLOGICAL GENETICS

— . (1934c). III. (Jber Genomwirkungen und Formbildungsfaehig-

keil bei Acetabularia. Arch. Enturicklgmech. 132.

(1934d). IV. Bntwicklungsphysiologische und genetische ( rrund-

lagen der Formbildung bei der Schirmalge Acetabularia. Naiurwiss.

22.

— . (1036). Studies zum Polaritaetsproblem. I III. Z<><>1. Jahrb.

56. Ahl.f. AUg. Znol. a. Physiol.
1 1 igbdoorn, A. L. (1911). Autokatalytic substances the determinants for

the inheritable characters. Vortr. Avf. Enturicklgmech. 12.

— . (1934). Die Mutation ini Liehte der biomechanischen Krhlich-

keitstheorie. Zeitsch. ind. Abstl. 67, 205-207.
Hagiwara, T. (1926). Genetic studies on Impatiens balsamin a. (.Japa-
nese, English summary.) Bot. Mag. Tokyo 40.
Haqiwaba, T. (1931). The genetics of flower colour in Pharbitis nil.

Jour. Coll. Agric. Tokyo 11.
. (1932). On the genetico-physiological studies of the colour

development of flowers in Pharbitis nil. Proc. Imp. Acad. Japan 8.
Haldane, J. B. S. (1930). A note on Fisher's theory of the origin of

dominance, and on a correlation between dominance and linkage.

Anier. Naturalist 64.
— . (1932). The time of action of genes, and its bearing on some

evolutionary problems. Amer. Naturalist 66, 5-24.

(1935). Les contributions de la genet ique a la solution de quelques

problemes physiologiques. Reunion pleniere de la SociiU de Biol.

Paris.
Hamburger, Viktor. (1936). The larval development of reciprocal

species hybrids of Triton taeniatus, Leyd. (and Triton palmatus, Duges)

X Triton cristatus, Laur. Jour. Exp. Zool. 73.
Ha.mmersohlag, V. (1932a). Uber die erstmalige Entstehung und die

Verschiedenheit des Erbganges hereditarer menschlicher und tierischer

Organopathien. Wiener klin. Woschr. No. 9.

— . (19326). Uber kombinierte Heredopathien und ihren mutmasz-

lichen Erbgang. Wiener klin. Woschr. No. 25 and 26.

— . (1932c). Ungemage Lebenslage als idiokinetischer Faktor.

Eine historisch-kritische Betrachtung. Zeitschr. f. Konstitutionsl.

16, 607-620.

— . (1932d). Die hereditar-degenerative Innenohrerkrankung (Here-

dopathia acustica) vom Standpunkt der Erbpathologie. Zugleich ein

Beit rag zur Lehre von der Polyallelie beim Menschen. Moschr.

Ohrenheilkd., Laryngo-Rhinol. 66, 1281-1519.
. (1933a). Uber Polyallelie bei Mensch und Tier und iiber das

Dominanzphanomen. I. Mitteilung. Zugleich ein Beitrag zu Gold-

schmidts Theorie der Vererbung. Zeitschr. Konstitutionsl. 17, 393-417.
. (19336). Uber Polyallelie und iiber das Dominanzphanomen.

II. Mitteilung. Die japanische Tanzmaus und die "Schuttelmaus"

von Lord und Gates. Zeitschr. Konstitutionsl. 17, 418-427.
. (1933c). Die hereditare Innenohrerkrankung im Lichte der

Vererbungslehre. Klin. Woschr. 12, 1903-1906.

BIBLIOGRAPHY 335

. (1935). t'ber die additive Wirkung homologer Allele. Genelicn

17, 270-290.
Hammett, F. S. (1931). The proliferation reaction of the skin to sulfhy-

dryl and its biological significance. Protoplasm 13.
Hanson, F. B. (1928). The effect of X-rays in producing return gene

mutations. Science 67, 562-563.

and F. Heys. (1929). An analysis of the effects of the different

rays of radium in producing lethal mutations in Drosophila. Amer.
Naturalist 63.

Harder, R. (1927). Zur Frage nach der Rolle von Kern und Protoplasma
im Zellgeschehen und bei der fjbertragung von Eigcnschaften. Zeitschr.
f. Botanik 19.

. (1934). t)ber die Musterbildung an Petunienbluten. Nachr.

Ges. Wiss. Gottingen Math.-Physik. Kl. Fachgr. VI, Biologie, N.F.
1, No. 2.

Hardesty, Mary. (1933). The feather of the guinea fowl and mathe-
matical theory of individual feather patterns. Jour. Exp. Zool. 66.

. (1935). The effect of thyroxin injections upon the feather of the

guinea fowl. Jour. Exp. Zool. 71, 389-416.

Harland, S. C. (1932). The genetics of Gossypium. Bibliog. Genetica 9.

. (1934). The genetics of cotton. IX. Further experiments on the

inheritance of the crinkled dwarf mutant of G. barbadense L. on inter-
specific crosses and their bearing on the Fisher theory of dominance.
Jour. Genetics 28.

Harnly, M. H. (1929). An experimental study of environmental factors
in selection and population. Jour. Exp. Zool. 53, 141-170.

. (1930a). A critical temperature for lengthening of the vestigial

wings of Drosophila melanogaster with sexually dimorphic effects.
Jour. Exp. Zool. 56, 363-368.

. (19306). A sexually dimorphic response to a critical temperature

during a critical period of development in Drosophila melanogaster.
Proc. 2d Internal. Congr. Sex Res.

. (1932). The temperature-effective period for the lengthening

of the vestigial wings of Drosophila. Proc. 6th Internal. Congr. Genetics 2.

. (1933). The growth curves of the length of the vestigial wings

of D. melanogaster. Amer. Naturalist 67.

. (1936). The temperature-effective periods and the growth

curves for length and area of the vestigial wings of Drosophila melano-
gaster. Genetics 21.

and M. L. Harnly. (1935). The effects of temperature on the

wings and the eyes of the dimorphous vestigial combination in D.
melanogaster. Jour. Exp. Zool. 72.

and . (1936). The effects of the gene on growth and

differentiation as shown by the temperature responses of pennant and
its heterozygote in D. melanogaster. Jour. Exp. Zool. 74.
Harrison, J. W. H. (1920a). Genetical studies in the moths of the
geometrid genus Oporabia (Oporinia) with a special consideration of
melanism in the lepidoptera. Jour. Genetics 9.

336 PHYSIOLOGICAL GENETICS

. (19206). The Inheritance of Melanism in the genus Tephrosia

(Ectropifl) with some consideration of 1 1 1 * - inconstancy of unit char-
acters under crossing. Jour. Genetics 10.
Harbison, R. (i. (1924). Some unexpected results of the heteroplastic

transplantation of limbs. Proc. Nat. Acad. Sci. 10.
— : . (1929). Correlation in the development and growth of the eye

studied by means of heteroplastic transplantation. Arch. Enivnckig-

mech. Abt. 120.

. (1937). Embryology and its relations. Science 86.

Hasimoto, H. (1930). Hereditary superfluous legs in the silkworm.

Jap. Jour. Genetics 6.
Heilbronn, A. (1933). tlber die genetische Legalisation Stoffwechsel

und Generationsfolge regvdierender Anlagen. Biol. Zentralbl. 63,

431-441.
Henke, K. (1924). Die Faerbung und Zeichnung der Feuerwanze

(Pyrrhocoris apterus L.) and ihre experimentelle Beeinflussbarkeit.

Zeitschr. vgl. Physiol. 1.
. (1928). tlber die Variabilitat des Fliigelmusters bei Larontia

sordidata F. und einigen anderen Schmetterlingen. Zeitschr. f. Morph.

u. Oekol. 12, 240-282.
. (1933a). Untersuchungen an Philosamia cynthia Drury zur

Entwicklungsphysiologie des Zeichnungsmusters auf dem Schmetter-

lingsfluegel. Arch. Entincklgmech. 128.
. (1933/;). Zur Morphologie und Entwicklungsphysiologie der

Tierzeichnungen. Naturwiss. 21.
. (1933c). Zur vergleichenden Morphologie des zentralen Sym-

metriesystems auf dem Schmetterlingsfluegel. Biol. Zentralbl. 63.
. (1935). Entwicklung und Bau tierischer Zeichnungsmuster.

Verh. Deutsch. Zool. Ges.
. (1936). Versuch einer vergleichenden Morphologie des Fluegel-

musters der Saturniden auf entwicklungsphysiologischer Grundlage.

Nova acta Leopoldina N.F. 4.
. (1937). Allgemeine Genetik (einschl. Genphysiologie). Fortschr.

Zool. N.F. 1.
Hersh, A. H. (19246). The effect of temperature upon the heterozygotes

in the bar series of Drosophila. Jour. Exp. Zool. 39.
. (1927). Temperature effects in reciprocal crosses of the bar series

of Drosophila. Jour. Exp. Zool. 47.
. (1928). Organic correlation and its modifications in the bar series

of Drosophila. Jour. Exp. Zool. 50, 239-255.
. (1929). The effect of different sections of the X-chromosome

upon bar eye in Drosophila melanogaster. Amer. Naturalist 63.
. (1930a). The facet-temperature relation in the bar series of

Drosophila. Jour. Exp. Zool. 57, 283-306.
. (19306). The need for and the proposal of a new genetic term.

Science 72, 294-295.
. (1931). Facet number and genetic growth constants in bar-eyed

stocks of Drosophila. Jour. Exp. Zool. 60, 213-248.

BIBLIOGRAPHY 337

— . (1932). The relative growth function in its application to the

individual and to the group. Proc. 6th Internat. Congr. Genetics 2.
— . (1934a). Evolutionary relative growth in the Titanotheres.

Amer. Naturalist 68.
— . (19346). On mendelian dominance and the serial order of

phenotypic effects in the bar series of Drosophila melanogaster. Amer.

Naturalist 68.
— . (1934c). The time curve of facet determination in an ultrabar

stock of Drosophila melanogaster. Jour. Gen. Physiol. 17.

and Esther Ward. (1932). The effect of temperature on wing

size in reciprocal heterozygotes of vestigial in Drosophila melanogaster.

Jour. Exp. Zool. 61, 223-244.
Hersh, R. K. (1924a). The effect of temperature upon the full-eyed race

of Drosophila. Jour. Exp. Zool. 39.
Hertweck, H. (1931). Anatomie und Variabilitaet des Nervensystems

und der Sinnesorgane von Drosophila melanogaster (Meigen). Zeitschr.

Wiss. Zool. 139.
Hertwig, Paula, and Tine Rittershaus. (1929). liber Fehlfedern bei

gesperberten Hiihnern. Arch. f. Geflugelkde. 3, H. 3, 65-76.
Hinrichs, M. A. (1925). Modification of development on the basis of

differential susceptibility to radiation I. Jour. Morphol. Physiol. Bd.

41.
Hirata, M. (1935). Physical investigation on the origin of coloured

patches of some kidney-beans, Phaseolus vulgaris L. Sci. Paper Inst.

Phys. Chem. Res. 26.
Hoerstadius, Sven. (1936). Studien liber Heterosperme Seeigelmerogone

nebst bemerkungen iiber einige Keimblattchimaeren. MusSe Roy.

Hist. Naturelle Belgique 2 S6r. Fasc. 3.
Hoge, M. A. (1915). The influence of temperature on the development

of a Mendelian character. Jour. Exp. Zool. 18.
Holmes, Anne. (1935). The pattern and symmetry of adult plumage

units in relation to the order and locus of the embryonic feather papillae.

Amer. Jour. Anat. 56, 512-535.
Holtfreter, J. (1936). Regionale Induct ionen in xenoplastisch zusam-

mengesetzten Explantaten. Arch. Entincklgmech. 134.
Honing, J. A. (1923). Nicotiana deformis n. sp. und die Enzymtheorie

der Erblichkeit. Genetics 5.
. (1927). Dominanzwechsel bei der Lichtkeimung. Verh. 5

Int. Kongr. Vererb. Bin. 1927. Zeitschr. ind. Absll. Suppl. Bd. 2, 861-

865, 1928.
. (1930). Nucleus and plasma in the heredity of the need of light

for germination in Nicotiana seeds. Genetica 12, 441-468.

(1931). Canna crosses. III. Plasmatic influence's. IV. Canna

aureo-vittata gigas a vegetative mutation. Mededeel. Landbouwhooge-

school Wageningen 35, Paper 1.
Houghtaling, Helen. (1935). A developmental analysis of size and

shape in tomato fruits. Bull. Torrey Bot. CI.
Huxley, J. S. (1932). Problems of relative growth. London.

."MS I'HYSIOUH.ICAL GENETICS

— . (1935a). Chemical regulation and the hormone concept. Biol,

Rev. 10.

— . (1935b). The field concepl in biology. Trans, on the Dynamic*

of l>< Vi lop. 10.

— . (1935c). The interaction of bar and riot-bar tissue in Drosophila

mosaic eyes: an embryological hypothesis. .Jam-. Hud. 4.

and A. Wolsky. (1932). Structure of normal and mutanl eves in

Gammarus chevreuxi. Nature Feb. 13.

and — . (1936). Zur Frage der entwicklungsphysiologischen

Determination des Arthropodenauges. Biol. Zentralbl. 56.

I

Iuiin, N. A. (1926). Studies in morphogenetics of animal pigmentation.

I. Morpliogenetic analysis of the genet ical constitution in albino guinea-
pigs. 1 1. Investigations of the temperature influence of 1 he Himalayan
rabbit's pigmentation. (Russian, English summary.) Trans. Lab.
Ex p. Biol. Zoopark Moscow 1, 1-63.

— . (1927). Analysis of the pigment formation by low temperature.
(Studies in morphogenetics of animal pigmentation IV.) (Russian,
English summary.) Trans. Lab. Exp. Biol. Zoopark Moscow 3, 183-
200.

and W. N. Iljin. (1929). Die Abhangigkeit der Farbung der

Siamesischen Katzen von der Temperatur. (Russian, German sum-
mary.) Trans. Lab. Exp. Biol. Zoopark Moscow 5, 217-234.

I mm, V. (1925). Genetic behaviour of the willow leaf in the Japanese
morning glory. Jour. Genetics 16.

. (1934). On the mutable genes of Pharbitis, with special reference

to their bearing on the mechanism of Bud- Variation. Jour. Coll.
Agric. Tokyo Imp. Univ. 12.

and D. Moriwaki. (1936). A probable case of cytoplasmic

inheritance in man: a critique of Leber's disease. Jour. Genetics 33.
Ives, P. T. (1935). The temperature-effective period of the scutc-1
phenotype. Proc. Nat. Acad. Sci. 21, 646-650.

Jt/HN, Mary. (1933). Individual feather succession in the hybrid capon.
I'roc. Soc. Exp. Biol. Med. 30.

— , G. H. Faulkner, and R. C. Gustavson. (1931). The correlation
of rates of growth and hormone threshold in the feathers of fowls.
Jour. Exp. Zool. 68.

- and R. M. Fraps. (1934). Pattern analysis in plumage. I.
Curve of barb growth. II. Methods of definitive feather analysis.
III. Action of thyroxin in high concentrations. IV. Order of asym-
metry in the breast tracts. Proc. Soc. Exp. Biol. Med. 31, 1181-
1190.

and— — . (1936). Developmental analysis in plumage. I. The

individual feather: methods. Physiol. Zool. 9.

BIBLIOGRAPHY 339

and R. G. Gustavson. (1930). The production of female

genital subsidiary characters and plumage sex characters by injection

of human placental hormone in fowls. Jour. Exp. Zool. 66.
Just, E. E. (1936). A single theory for the physiology of development and

genetics. Amer. Naturalist 70.
Johannsen, O. A. (1924). Eye-structure in normal and eye-mutant

Drosophilas. Jour. Morph. Physiol. 39.
Johannsen, W. (1909). Elemente der exakten Erblichkeitslehre. Jena.
Jollos, V. (1921). Experimented Protistenstudien I. Arch. Protiat< nk.

43.
. (1924). Untersuchungen iiber Vererbung und Variabilitaet bei

Arcellen. Arch. Protistenk. 49.
. (1930). Studien zum Evolutionsproblem. I. Uber die experi-

mentelle Hervorrufung und Steigerung von Mutationen bei Drosophila

melanogaster. Biol. Zentralbl. 50.
. (1933). Die Ubereinstimmung der bei Drosophila melanogaster

nach Hitzeeinwirkung entstehenden Modifikationen und Mutationen.

Naturiviss. 21, 831-834.

(1934). Inherited changes produced by heat-treatment in

Drosophila melanogaster. Genetica 16.
Jones, D. F. (1935). Somatic segregation due to hemizygous and missing

genes and its bearing on the problem of atypical growth. Proc. Nat.

Acad. Sci. 21, 90-96.
Jucci, C. (1928). Come si eredita la capacita di accrescimento negli

incroci reciproci tra le due razze di bachi da seta (Bombyx mori) Varo e

Bianco cinese. Boll. Soc. Natur. Napoli 40, 163-195.
. (1929). L'eredita citoplasmatica. (Prelezione). Studi Sassaresi

Ser. 2, 7, 3-17.

K

Kaiser, Samuel. (1935). The factors governing shape and size in Capsi-
cum fruits; a genetic and developmental analysis. Bull. Torrey Bot. CI.
62.

Kamenoff, R. J. (1933). Effects of the flexed-tailed gene on the develop-
ment of the house-mouse. Jour. Morph. 58.

. (1934). The development of flexed tail in the mouse. Disserta-
tion, Columbia University.

Karczag, L. (1928). Die Stereogene als Erbeinheiten. Eine neue
Theorie der Vererbungserscheinungen. Zeitschr. ind. Abstl. 48, 86-144.

Karrer, P. and E. Helfenstein. (1932). Plant pigments. Amer. Rev.
Biochem. 1.

Keller, Caecilie. (1933). Vergleichende Zellen- und Kernmessungen
bei grossen und kleinen Huhnerrassen zur Priifung der genetisch beding-
ten Wuchsunterschiede. Zugleich ein Beitrag zur Frage des rhyth-
mischen Wachstums der Kerne. Zeitschr. Zellforsch. 19, 510-536.

Kellicott, W. E. (1916). The effects of low temperature upon the
development of Fundulus. Amer. Jour. Anat. 20.

340 PHYSIOLOG1CM. GENETICS

Kellogg, V. L. (1908). Inheritance in Silkworms, I. I.eland-Standford

Junior University Publications. University Series.
Kkkkis, .1. (1U33). Einfluss der Temperatur auf die Entwicklung der

Hybriden von Drosophila melanogaster X Drosophila Bimulans.

Arch. Entvnoklgmech. 130, 1 10.
Kikkawa, II. (1934). A consideration on the gene-manifestation.

(Japanese, English resume- p. 2014. ) Jap. Jour. Bot. Zool. 2, '2009-2014.
Kin.;, Helen D. (1931). Studies on the inheritance of structural anoma-
lies. A im i. Jour. Anat. 48, 231-260.
Kniesche, G. (1914). Uber die Farben der Vogelfedern. I. Die

Gruenfaerbung auf Grundlage der Blaustruktur. Zool. Jahrb. {Anat.).

38.
Kobozieff, N. (1935). Recherches morphologiques et genetiques sur

l'anourie chez la souris. Bull. Biol. 69.
Koehler, W. (1932). Die Entwicklung der Fluegel bei der Mehlmotte

Ephestia kuehniella Zeller mit besonderer Beruecksichtigung des

Zeichnungsmusters. Zeitschr. f. Morph. u. Oekol. 24.
. (1934). Die Entwicklung der Fluegel bei der Mehlmotte Ephestia

kuehniella Zeller, mit besonderer Beruecksichtigung des Zeichnungs-
musters. Zeitschr. f. Morph. u. Oekol. Tiere. 24.
Kohler, W., and W. Feldotto. (1935). Experimentelle Untersuchungen

liber die Modifikabilitat der Fliigelzeichnung, ihrer Systeme und

Elemente in den sensiblen Perioden von Vanessa urticae L., nebst

einigen Beobachtungen an Vanessa io L. Arch. Jul. Klaus-Stiftg.

10, 313-453.
Kofoid, C. A. (1895). On the early development of Limax. Bull. Mus.

Harvard Coll. 27.
Koller, P. (1930). On pigment formation in the D-black rabbit. Jour.

Genetics 22, 103-107.
Koltzoff, N. K. (1928). Physikalisch-chemische Grundlage der Morpho-

logie. Biol. Zentralbl. 48.
. (1934). Genetics and physiology of development. (Russian.

English summary). Russ. Zeitschr. f. Biologie 3, 420-456.

— . (1935a). Physiologie du developpement et genetique. Actual.

Scientif. Indust. 254. Hermann et Cie. Paris.

(19356). The role of the gene in the physiology of development.

(Russian. English summary, p. 773-774). Zeitschr. f. Biol. (Russian)

4, 753-774.
Komai, T. (1926). Crippled, a new mutant character of Drosophila

melanogaster and its inheritance. Genetics 11.
. (1927). Genetic studies on crippled, a mutant character of

Drosophila melanogaster. Mem. Coll. Sci. Kyoto Imp. Univ. Ser. B, 2.
Kopec, St. (1922). Studies on the necessity of the brain for the inception

of insect Metamorphosis. Biol. Bull. 42.
. (1927). A case of spotting in a Himalayan rabbit and its probable

explanation. Biologia Generalis 3, 411-418.
Kosminsky, P. (1909). Einwirkung aeusserer Einfluesse auf Schmetter-

linge. Zool. Jahrb. (Abt. system.) 27.

BIBLIOGRAPHY 341

— . (1924). Der Gynandrmnorphismus bei Lymantria dispar L.
unter der Einwirkung aeusserer Einfluesse. Vorl. Mittlg. Biol.
Zentralbl. 44.

(1927). Intersexualitaet im maennlichen Kopulationsapparat

von Lymantria dispar L. unterm Einfluss der Temperatur. Biol.

Zentralbl. 47.
Kosswig, C. (1927). Uber die Vererbung und Bildung von Pigment bei

Kaninchenrassen. Zeitschr. ind. Abstl. 46, 368-401.
. (1929a). Uber die veranderte Wirkung von Farbgenen des

Platypoecilus in der Gattungskreuzung init Xiphophorus. Zeitschr.

ind. Abstl. 60, 63-73.

(19296). Zur Frage der Geschwulstbildung bei Gattungsbas-

tarden der Zahnkarpfen Xiphophorus und Platypoecilus. Zeitschr.
ind. Abstl. 52, 114-120.

Kostoff, D. (1929). An androgenic Nicotiana haploid. Zeitschr. f.
Zellforsch. 9.

. (1935). On the increase of mutation frequency following inter-
specific hybridization in Nicotiana. Cur. Sci. 3.

Krafka, Jr., J. (1920a). The effect of temperature upon facet number in
the bar-eyed mutant of Drosophila. Jour. Gen. Physiol. 2.

. (19206). Environmental factors other than temperature affect-
ing facet number in the bar-eyed mutant of Drosophila. Jour.
Gen. Physiol. 3.

(1924). Development of the compound eye of Drosophila melano-

gaster and its bar-eyed mutants. Biol. Bull. 47.
Krieg, H. (1922). Streifung und Stromung. Versuch einer kausalen

Begriffsbestimmung einiger Grundelemente der Saeugetierzeichnung.

Arch. Entwicklgmech. 61.
Kroning, F., and C. Engelmann. (1934). Penetranz, Expressivitat und

Asymmetrie der tlberzehen des Meerschweinchens und die Abhangig-

keit ihrer Realisation von Aussenfaktoren. Nachr. Ges. Wiss. Gottingen

Math.-Physik. Kl. Fachgr. VI N.F. 1, 123-141.
Krueger, A. B. (1936). The nature of bacteriophage and its mode of

action. Physiol. Rev. 16.
Ktjehn, A. (1926). tjber die Aenderung des Zeichnungmusters von

Schmetterlingen durch Temperaturreize und das Grundschema der

Nymphalidenzeichnung. Nachr. Ges. Wiss. Goettingen, Math.-Phys. Kl.
. (1927). Die Pigmentierung von Habrobracon juglandis Ashmed,

ihre Praedetermination und ihre Vererbung durch Gene und Plasmon.

Nachr. Ges. Wiss. Goettingen, Math.-Phys. Kl.
. (1932a). Entwicklungsphysiologische Wirkungen einiger Gene

von Ephestia kuhniella. Natururiss. 20, 974-977.
. (19326). Zur Genetik und Entwicklungsphysiologie des Zeich-

nungsmusters der Schmetterlinge. Nachr. Ges. Wiss. Gottingen

Math.-Physik. Kl. Fachgr. VI, 312-335.
. (1936). Weitere Untersuchungen ueber den Gen-A-Wirkstoff

bei der Mehlmotte Ephestia kuehniella Z. Nachr. Ges. Wiss. Goettingen

Math.-Phys. Kl.

342 PHYSIOLOGICAL GENETICS

. I.. Caspabi, and E. Plagge. (1935). Uber hormonale ( lenwirkun-

gen bei Epheetia kuehniella Z. Nackr. Gee. Wise. GoeUingen Math.
Phys. Kl.

— and M. v. Engelhardt. (1933). Ober die Determination dea
Symmetriesystems auf dem Vorderfluegel von Ephestia kuehniella
Zeller. Arch. Entwicklgmech. 130.

— and M. v. Engelhardt. (1936). Uber die Determination dee
Fluegelmusters bei Abraxas gros8ulariata L. Nachr. Ges. Wis*.
GoeUingen, Math -Phys. Kl. 2.

— and K. Henke. (1929). Genetische und entwicklungsphysio-
Logische Untersuehungen an der Mchlinotte Ephestia kuhniella Zeller.
I-VII. Ahhandl. Ges. Wiss. Gottingen Math.-Phys. Kl. A'./'. 16, 1-121.

and . (1930). Eine Mutation der Augenfarbe und der

Entwicklungsgeschwindigkeit bei der Mehlmotte Ephestia kuhniella Z.

Arch. Entwicklgmech. 122, 204-212.
and . (1932). Genetische und entwicklungsphysiologische

Untersuehungen an der Mehlmotte Ephestia kuhniella Zeller. VI I U

XII. Abhandl. Ges. Wise. Gottingen Math.-Phys. Kl. N.F. 15, 127-219
and . (1935). Uber einen Fall von geschlechtsgekoppolter

Vererbung mit wechselnder Merknialsauspraegung bei der Mehlmotte

Ephestia kuehniella Zeller. Ibid.
ancl . (1936). Genetische und Entwicklungsphysiologische

Untersuehungen an der Mehlmotte Ephestia kuehniella Zeller I-XIV.

Abh. Ges. Wiss. GoeUingen. Math.-Phys. Kl. N.F. 15.

and H. Piepho. (1936). Uber hormonale Wirkungen bei der

Verpuppung der Schmetterlinge. Nachr. Ges. Wiss. GoeUingen Math.-
Phys. Kl. N. F. 15.

Kuester, E. (1913). Uber Zonenbildung in kolloiden Medien. Jena.

Kuhn, O. (1928). Uber die Pigmentierung des Gefieders bei Htihner-
bastarden. (Vorl. Mitt.) Zeitschr. f. Ziichtungskde 3, 615-619.

. (1932). Entwicklungsphysiologische Untersuehungen an der

Vogelfeder. Arch. Entwicklgmech. 127.

and J. Kuhn. (1926). Uber die Scheckung der Haustaube.

Biol. Zentralbl. 46.
Kuntze, R., and G. Poluszynski. (1929). Genetische Analyse der
Farbungsvariabilitat des Kiefernspinners (Dendrolimus pini L.).
(Lepidoptera.) Bull. Acad. Pol. Sci. Leltr. CI. Sci. Mathim. natur. Ser.
B, 153-189.

Lancefield, D. E. (1918a). An autosomal bristle modifier, affecting a
sex-linked character. Amer. Naturalist 52.

. (19186). Scarlet, an autosomal eye color identical with sex-
linked vermilion. Biol. Bull. 35.

Landauer, W. (1928). The morphology of intermediate rumplessness
in the fowl. Jour. Heredity 19.

. (1931). Untersuehungen uber das Kruperhuhn II. Zeitschr.

mikr. an. Forsch. 25.

BIBLIOGRAPHY 343

— . (1932). Studies on the creeper fowl. III. The early development

and lethal expression of homozygous creeper embryos. Jour. Genetics

25, 367-394.
— . (1933). A gene modifying frizzling in the fowl. Jour. Heredity

24, 153-156.
. (1934). Studies on the creeper fowl VI. Bull. Starrs Exp. Sta.

193.
— and L. C. Dunn. (1930a). The "Frizzle" character of fowls.

Its expression and inheritance. Jour. Heredity 21, 291-305.

and . (19306). Studies on the creeper fowl. I. Genetics.

Jour. Genetics 23, 397-413.
Lang, A. (1908). tjber die Bastarde von Helix hortensis Mueller und
Helix nemoralis L. (Mit Beitraegen von Bosshard, Hesse und Kleiner).
Festschrift d. Universitaet Jena.
Lawrence, W. J. C. (1931). The genetics and cytology of Dahlia
variabilis. Mutation or segregation in the octoploid Dahlia variabilis.
Jour. Genetics 24, 257-324.

— and Rose Scott-Moncrieff. (1935). The genetics and chemistry
of flower colour in dahlia; a new theory of specific pigmentation.
Jour. Genetics 30, 155-226.
Lebedeff, G. A. (1932). Interaction of ruffled and rounded genes of
Drosophila virilis. Proc. Nat. Acad. Sci. 18, 343-349.

. (1933). Studies on factor interaction in Drosophila virilis.

Amer. Naturalist 67.

. (1934). Genetical hermaphroditism in Drosophila virilis. Proc.

Nat. Acad. Sci. 20.

(1935). Further studies on factor interaction in Drosophila

virilis. Genetics 20.
Lehmann, F. E. (1936). Stehen die Erscheinungen der Otocephalie und

der Cyclopie bei Triton mit Axialgradienten oder mit Stoerungen

bestimmter Organisatorregionen im Zusammenhang? Rev. Suisse

Zool. 43.
Lesley, Margaret M., and H. B. Frost. (1927). Mendelian inheritance

of chromosome shape in Matthiola. Genetics 12, 449-460.
Lewit, S. G. (1930a). Step allelomorphism in Drosophila melanogaster.

V. Jour. Exp. Biol. 6.
. (19306). Untersuchungen liber Treppenallelomorphismus. V.

Die Mutation scute-9 bei der Drosophila melanogaster und die Frage

nach den Allelomorphen-Analysatoren. Arch. Entwicklgmech. 122,

770-783.
Li, J. C. (1927). The effect of chromosome aberrations on development

in Drosophila melanogaster. Genetics 12, 1-58.
Lilienfeld, F. A. (1929). Vererbungsversuche mit schlitzblaettrigen

sippen von Malva parviflora. Biblioth. Genet. 13.
Lillie, F. R. (1927). The gene and the ontogenetic process. Science

66, 361-368.
. (1929). Embryonic segregation and its role in the life history.

Arch. Entwicklgmech. 118, 499-533.

;; I I PHYSIOLOGICAL GENETICS

and Mart Juhn. (1032). The physiology <>f development of

Feathers. I. Growth-rate and pattern in the individual feather.

Physiol. Zool. 6.
Lindabx, Per Eric, (l'.o-ttii. Zur Kenntnis der physiologischerj Grund-

lagerj der Determination im Seeigelkeim. Ada Zool. 17.
Linden, M. v. (1899). Versuche iiber den Einfluss aeusserer

Verhaeltnisse auf die Gestaltung der Schmetterlinge. Illustr. Zcitschr.

f. Entomol.
. (1904). Die Ergebnisse der experiinentellen Lepidopterologie.

Biol. Zentralbl. 24.
Lindstrom, E. W. (1928). Linkage of size, shape, and colour genes in

Lycopersicum. Zeitschr. ind. Abstamm.-u. Vererb. Lehre, Supplbd. 2.

— . (1935). Segregation of quantitative genes in tetraploid tomato

hybrids as evidence for dominance relations of size characters. Genetics

20, 111.

and F. Gerhardt. (1926). Inheritance of carbohydrates and fat

in crosses of dent and sweet corn. Agric. Exp. Sta. Iowa State Coll.

Agri. Mech. Arts. Bull. 98.

— and . (1927). Inheritance of chemical characters in maize.

Iowa Slate Coll. Jour. Sci.
Little, C. C. (1913). Experimental studies of the inheritance of color

in mice. Carnegie Inst. Washingston Publ. 179.
. (1919). The fate of individuals homozygous for certain colour

factors in mice. Amer. Naturalist 53.
and H. J. Bagg. (1924). The occurrence of four inheritable

morphological variations in mice and their possible relation to treatment

with X-rays. Jour. Exp. Zool. 41.

and . (1929). The occurrence of four inheritable morpho-

logical variations in mice etc. Jour. Exp. Zool. 41.

Loeb, J. (1915). The blindness of the cave fauna etc. Biol. Bull. 27.

Luce, W. M. (1926). The effect of temperature on infrabar, an allelo-
morph of bar-eye in Drosophila. Jour. Exp. Zool. 46.
— . (1931). The temperature-effective period in infra-Bar and its
heterozygotes. Jour. Exp. Zool. 69.
— . (1935). Temperature studies on bar-infrabar. Jour. Exp. Zool.

71.
Ludwig, W. (1932). Das Rechts-Links- Problem in Tierreich und beim

Menschen. Berlin. I. Springer.
. (1936). Bestimmung und Vererbnng der Asyminetrieform

(Rechts-Links-Problem). Verhandl. Deutsch. Zool. Ges. 38. Zool.

Anz. Suppl. 9.
Lynch, C. J. (1919). An analysis of certain cases of intra-specinc sterility.

Genetics 4.

M

Machida, J. (1924). Crossing experiments with gipsy moths. Jour.
Coll. Agric. Tokyo 7.

BIBLIOGRAPHY 345

McClendon, J. J. (1912). The effects of alkaloids on the development of

fish (Fundulus) eggs. Amer. Jour. Phys. 31.
MacC'racken, J. (1909). Heredity of the race-characters univoltinism

and bivoltinism in the silkworm (Bombyx mori). A case of non-

Mendelian inheritance. Jour. Exp. Zool.
Mangelsdorf, P. C, and G. S. Fraps. (1931). A direct quantitative

relationship between vitamin A in corn and the number of genes for

yellow pigmentation. Science 73, 241-242.
Margolis, O. S. (1934). The effect of a supernumerary X chromosome

on members of the bar series of Drosophila. Genetics 19, 40-61.
. (1935a). Studies on the Bar series of Drosophila. I. The effect

of the gene vestigial on facet number in Bar. Genetics 20, 156-171.
— . (19356). Studies on the Bar series of Drosophila melanogaster.

II. The temperature-effective period for facet determination in Bar

und Bar vestigial. Genetics 20, 207-222.
. (1936). Studies on the bar series of Drosophila. III. The facet

relations in dorsal and ventral lobes of the eye. Genetics 21.

and C. W. Robertson. (1937). Studies on the Bar series of Droso-

phila. IV. The temperature-effective period for facet determination

in the wild type. Genetics 22.
Martini, E. (1923). Die Zellkonstanz und ihre Beziehungen zu anderen

zoologischen Vorwuerfen. Zeitschr. f. Ges. Anat. Bd. 70.
Marza, V. (1929). Le problemederontogenesechimiqueetl'historeaction

de la lecithine pendant revolution du systeme nerveux de l'embryon de

poule. Cornpt. Rend. Soc. (roumaine) Biol. 102, 195.
Medvedev, N. N. (1935). Genes and development of characters. I.

The study of the growth of the imaginal discs of eyes of the wild type

larvae and three mutants — Lobec, glass2 and eyeless2 in Drosophila

melanogaster. Zeitschr. ind. Abetl. 70, 55-72.
. (1937). Body-color mutants in Drosophila as studied by trans-
plantation. Compt. Rend. Acad. Sci. USSR 14.
Melchers, G. (1935). Uber reziprok verschiedene Merkmalsausbildung

in der Fi der Kreuzung Saxifraga ascendens L. X S. tridactylites L.

unter Beriicksichtigung des Entwicklungsstadiums. Zeitschr. ind.

Abstl. 69, 263-277.
Merrifield, F. (1889). Temperature experiments on Lepidoptera.

Proc. Entomol. Soc, London.
. (1894). Experiments in temperature-variation on Lepidoptera,

and their bearing on theories of heredity. Proc. Entomol. Soc, London.
Michaelis, P. (1933). Entwicklungsgeschichtlich-genetische Untersuch-

ungen an Epilobium. II. Die Bedeutung des Plasmas fuer die Pollen-

fertilitaet des Epilobium luteum-hirsutum-Bastardes. Zeitschr. f.

ind. Abstl. 66.
. (1934). Weitere Untersuchungen liber das Problem der Plas-

mavererbung. Der Ziichter. 6, 303-306.
. (1935a). Entwicklungsgeschichtlich-genetische Untersuchungen

an Epilobium. III. Zur Frage der Uebertragung von Pollenschlaueh-

346 PHYSIOLOGICAL GENETICS

plasma in die Eiselle und ihre Bedeutung fuer die Plasmavererbung.

Planta 23.
. (19356 . EntwicklungBgeschiclitlich-ncnctisclio Untersuchungen

an Epilobium. [V. Der Einfluss des Plasmona auf Verzweigung und

Pilzresistrux. Ber. Deutsch. Bot. Ges. 63.
. (1935c). Erhohte Wachstumaintensit&l und Pilzreaiatenz durcfa

Plasmavererbung, Bowie liber die Bedeutung des Plasmas bei Kreuzungs-

Bchwierigkeiten. Der Ziichter. 7, 74-77.
and M. v. Bellingshausen. (1935). Entwicklungsgeschichtlich-

genetische (Jntersuchungen an Epilobium. VII. Experimentelle lTnter-

Buchungea ueber die Beeinflussung der Pollenfertilitaet unter besonderer

Beruecksichtigung der Plasmonwirkung. Jahrb. f. Wiss. Bot. 82.

and E. Wertz. (1935). Entwicklungsgeschichtbch-genetische

Untersuchungen an Epilobium. VI. Vergleichende Untersuchungen

ueber das Plasmon von Epilobium hirsutum, E. luteum, E. inontanum

and E. roseuin. Zeitschr. ind. Absll. 70.
Minami, S. (1925). Untersuchungeu uber das Fluegelmosaik intersexuel-

ler Maennchen von Lymantria dispar L. Arch. Entwicklgmech. 105.
Mittasch, A. (1935). Uber katalytische Verursachung im physiolo-

gischen Geschehen. (Auch ein Berzelius-Gedenken.) Naturwiss.

23, 361-383.

— . (1936). Uber Katalyse und Katalysatoren in Chemie und

Biologie. Naturwiss. 24.
Mohr, O. L. (1923a). Das Deficiency-Phaenomen bei Drosophila

melanogaster. Zeitschr. f. ind. Abstammungs-u. Vererb. Bd. 30.
. (19236). A genetic and cytological analysis of a section deficiency

etc. Zeitschr. f. ind. Abstammungs-u. Vererb. Bd. 32.
. (1929a). Exaggeration and inhibition phenomena encountered

in the analysis of an autosomal dominant. Zeitschr. ind. Absll. 50,

113-200.

. (19296). Letalfaktoren bei Haustieren. Z uechtungskunde 4.

. (1932). On the potency of mutant genes and wild-type allelo-
morphs. Proc. 6th Internal. Congr. Genetics 1, 190-212.

and Chr. Wriedt. (1919). A new type of hereditary brachy-

phalangy in man. Carnegie Inst. Washington Publ. 295.
Montalenti, G. (1933). L'ontogenesi degli ibridi fra Bufo vulgaris e

Bufo viridis. Physiol. Zool. 6.
. (1934a). L'azione degli ormoni sul piumaggio degli uccelli.

Rassegna di Clinica, Terapia e Sci. affini. 33, 5-23.

(19346). A physiological analysis of the barred pattern in

Plymouth Rock feathers. Jour. Exp. Zool. 69, 269-345.
Morgan, L. V. (1931). Proof that bar changes to not-bar by unequal

crossing-over. Proc. Nat. Acad. Sci. 17, 270-272.
Morgan, T. H. (1912). The masking of a Mendelian result by the

influence of the environment. Proc. Soc. Exp. Biol. Med. 9.
. (1915a). The infertility of rudimentary winged females of

Drosophila ampelophila. Amer. Naturalist 49.
. (19156). The role of the environment in the realization of a sex-

linked Mendelian character in Drosophila. Amer. Naturalist 49.

BIBLIOGRAPHY 347

— . (1919). The genetic and the operative evidence relating to
secondary sexual characters. Carnegie Inst. Washington (I'ubl. 285).

— and C. B. Bridges. (1919). The origin of gynandromorphs.
Carnegie Inst. Washington Publ. 278.

C. B. Bridges, A. H. Sturtevant. (1925). The genetics of

Drosophila. Bibliog. Genetica II.

Muentzing, A. (1932). Cyto-genetic investigations on synthetic Galeop-
sis Tetrahit. Hereditas 16.

. (1936). The evolutionary significance of allopolyploidy. Her-
editas 21.

Muller, H. J. (1922). Variations due to change in the individual gene.
Amer. Naturalist 66.

. (1928a). The measurement of gene mutation rate in Drosophila,

its high variability, and its dependence upon temperature. Genetics 13.

. (19286). The problem of genie modification. Zeitschr. ind.

Abstamm.-u. Vererb. Supl. 1.

. (1929). The gene as the basis of life. Proc. Internal. Congr.

Plant Sci. 1, 897-921.

. (1930). Types of visible variations induced by X-rays in Droso-
phila. Jour. Genetics 22, 299-334.

=— . (1932a). Further studies on the nature and causes of gene muta-
tions. Proc. 6lh Internal. Congr. Genetics 1.

. (19326). A viable two-gene deficiency phaenotypically resembling

the corresponding hypomorphic mutations. Jour. Heredity 26.

. (1936). Bar duplication. Science. 83.

, B. B. League, and C. A. Offermann. (1932). Effects of dosage

changes of sex-linked genes, and the compensatory effects of other
gene differences between male and female. Anat. Rec. 61.

and A. Prokofyewa. (1934). Continuity and discontinuity of the

hereditary material. Compt. Rend. Acad. Sci. USSR. 4.

and . (1935). The individual gene in relation to the chromo-

mere and the chromosome. Proc. Nat. Acad. Sci. 21.

, and D. Raffel. (1935). Minute intergenic rearrangement

as a cause of apparent "gene mutation"? Nature 135 I.

Muller, H. J., A. A. Prokofieva-Belgowskaja, and K. V. Kossikov.
(1936). Unequal crossing over in the Bar mutant -as a result of duplica-
tion of a minute chromosome section. Compt. Rend. Acad. Sci.
USSR. 2.

Murray, W. S., and C. C. Little. (1935a). Further data on the existence
of extra chromosomal influence on the incidence of mammary tumors in
mice. Science II.

and . (19356). The genetics of mammary tumor incidence

in mice. Genetics 20.

N

Nadler, J. E. (1926). Effects of temperature on length of vestigial wing

of Drosophila virilis. Genetics 11.
Navaschin, M. (1927). Ein Fall von Merogonie infolge Artkreuzung bei

Kompositen. Ber. Deutsch. Bot. Ges. 45.

348 PHYSIOLOGICAL GENETICS

. (1930). Quantity of hereditary material and expression of specific

characters. (Preliminary note.) Zeitschr. ind. Abatl. 66, 348-352.
Newman, II. H. (1909). The process of heredity as exhibited by the

development of Fundulus hybrids. Jour. Exp. Zool. 6.
. (1910). Further studies on the process of heredity in Fundulus

hybrids. Jour. Exp. Zool. 8.
. (1914). Modes of inheritance in teleost hybrids. Jour. Exp. Zool.

16.

(1916). Heredity and organic symmetry in armadillo quadruplets.

II. Mode of inheritance of double scutes and a discussion of organic

symmetry. Biol. Bull. 30.
Nilsson-Ehle, H. (1909). Kreuzungsuntersuchungen an Hafer und

Weizcn. Act. Univers. Lund.
Noujdin, N. I. (1936). Genetic analysis of certain problems of the

physiology of development of Drosophila melanogaster. Inst. Gen.

Acad. Set., Moscow 6.

0

Ohlkers, F. (1930). Studien zum Problem der Polymerie und des
multiplen Allelomorphismus II. Zeitschr. f. Botanik 23, 967-1003.

. (1935). Die Erblichkeit der Sepalodie bei Oenothera und

Epilobium. Studien zum Problem der Polymerie und des multiplen
Allelomorphismus. III. Zeitschr. f. Botanik 28, 161-222.
. (1936). Untersuchungen zur Physiologie der Meiosis III.

Zeitschr. f. Botanik 30.
Offermann, C. A. (1935). The position effect and its bearing on genetics.

Bull. Acad. Sci. USSR. CI. Sci. Math. Nat. 1.
Ogura, S. (1932). Erblichkeitsstudien am Seidenspinner Bombyx mori L.

III. Genetische Untersuchung der Haeutung (1. Teil). Z. ind. Abstl.

61.
Onslow, H. (1915). A contribution to our knowledge of the chemistry

of coat-colour in animals and of dominant and recessive whiteness.

Proc. Roy. Soc. London B 89.
Onslow, M. Wheldale. (1925). The Anthocyanin pigments of plants.

2d ed., Cambridge.
Overbeek, J. Van. (1935). The growth hormone and the dwarf type of

growth in corn. Proc. Nat. Acad. Sci. 21.

Panshin, J. B. (1935). New evidence for the position effect hypothesis.
Compt. Rend. Acad. Sci. USSR. N.S. 4.

Pariser, K. (1927). Die Zytologie and Morphologie der triploiden
Intersexe des rueckgekreuzten Bastards von Saturnia pavonia und
Saturnia pyri. Zeitschr. Zellf. 5.

Patterson, J. T. (1928). The effects of X-rays in producing mutations
in the somatic cells of Drosophila. Science 68.

. (1929a). The production of mutations in somatic cells of Droso-
phila melanogaster by means of X-rays. Jour. Exp. Zool. 53, 327-372.

BIBLIOGRAPHY 349

— . (19296). X-rays and somatic mutations. Jour. Heredity 20,

261-267.
— . (1930a). Proof that the entire chromosome is not eliminated in

the production of somatic variations by X-rays in Drosophila. Genetics

16, 141-149.
— . (19306). Somatic segregation produced by X-rays in Drosophila

melanogaster. Proc. Nat. Acad. Set. 16, 109-111.
- — . (1931). A mutable miniature gene in Drosophila melanogaster.

Amer. Naturalist 66.

(1933). The mechanism of mosaic formation in Drosophila.

Genetics 18.
Pearsall, W. H. (1927). On the relative size of growing plant organs.

Ann. Bot. 41.
Pernitzsch, F. (1913). Zur Analyse der Rassenmerkmale der Axolotl.

I. Die Pigmentierung junger Larven. Arch. Mikrosk. Anal. 82.
Philiptschenko, J. (1929). Gene und Entwicklung der Ahrenform beim

Weizen. Biol. Zentralbl. 49, 1-16.
Piepho, H. (1935). Uber die Temperaturmodifikabilitaet und Genetik

zweier rotaeugiger Rassen der Mehlmotte Ephestia kuehniella Zeller.

Arch. Entwicklgmech. 133.
Plagens, G. M. (1933). An embryological study of a special strain of

deformed X-rayed mice, with special reference to the etiology and

morphogenesis of the abnormalities. Jour. Morph. 65, 151-178.
Plagge, E. (1935). Die Pigmentierung der Imaginal- und Raupenaugen

der Mehlmotte Ephestia kuehniella Zeller bei verschiedenen Rassen,

Transplantattraegern und Rassenkreuzungen. Arch. Entwicklgmech.

132.
. (1936a). Bewirkung der Augenausfaerbung der rotaeugigen

Rasse von Ephestia kuehniella durch Implantation artfremder Hoden.

Nachr. Ges. Wiss. Goeitingen Math.-Phys. Kl.
. (19366). Der zeitliche Verlauf der Ausloesbarkeit von Hoden-

und Imaginalaugenfaerbung durch den A-Wirkstoff bei Ephestia und

die zur Ausscheidung einer wirksamen Menge noetige Zeitdauer.

Zeitschr. ind. Abstl.-u. Vererb. 72.

(1936c). Transplantationen von Augenimaginalscheiben zwis-

chen der schwarz- und der rotaeugigen Rasse von Ephestia kuehniella

Z. Biol. Zentralbl. 56.
Plough, H. H. and C. F. Holthausen. (1937). A case of high mutation

frequency without environmental change. Amer. Nat. 71.
Pltjnkett, C. R. (1926). The interaction of genetic and environmental

factors in development. Jour. Exp. Zool. 46.
. (1932). Temperature as a tool of research in phenogenetics:

methods and results. Proc. 6th Internal. Congr. Genet. 2.
Pol. E. (1921). Brachydactyly etc. Virchows Arch. f. Pathol. Anat. u.

Physiol. 229.
Poppelbaum, H. (1914). Studien an gynandromorphen Schmetterlings-

bastarden aus der Kreuzung von Lymantria dispar L. mit japonica

Motsch. Zeitschr. ind. Abstl. 11.

350 PHYSIOLOGICAL GENETICS

Powsnkk, \j. (1935). The effects of temperature on the duration of the

developmental Btages <>f Drosophila melanogaster. Physiol, Zool. 8,

474-520.
I'h/ihuwi, II. (1001). Experimentelle Studien ueber Regeneration.

Arch. Entwicklgmech. 11.

— . (1902). Experimentelle Studien ueber Regeneration: II. Arch.

Entwicklgmech. 13.

(1908). Vercrbungsversuche ueber asymmetrische Augenfarbe

bei Angorakatzen. Arch. Entuncklgmech. 20.
Pugh, C. E. M. (1933). Tyrosinase from the skin of certain black

rabbits. Bioch. Jour. 27.
Punnett, R. C, and P. G. Bailey. (1921). Genetic studies in poultry.

III. Henfeathered cocks. Jour. Genetics 11.

R

R VSMUSSON, J. (1933). A contribution to the theory of quantitative

character inheritance. Hereditas 18.
Redfield, Helen. (1926). The maternal inheritance of a sex-limited

lethal effect in Drosophila melanogaster. Genetics 11.
Reinig, W. F. (1928). Uber das Manifestieren zweier Genovariationen

bei Drosophila funebris. Biol. Zentralbl. 48, 115-125.
. (1930). Uber das Manifestieren zweier Genovariationen bei

Drosophila funebris. Arch. Entwicklgmech. 121, 1-38.
Renner, O. (1920). Mendelsche Spaltung und chemisches Gleichgewicht.

Biol. Centrbl. 40.
. (1922). Eiplasma und Pollenschlauchplasma als Vererbungs-

traeger bei den Oenotheren. Zeitschr. f. Abstl. u. Vererb. 27.
. (1924). Die Scheckung der Oenotherenbastarde. Biol. Zentralbl.

44.

— . (1925). Untersuchungen ueber die faktorielle ^Constitution

einiger komplexheterozygotischer Oenotheren. Bibliotheca Genet., Heft

9. Berlin.

— . (1929). Artbastarde bei Pflanzen. Handb. d. Ycrerbungsunssen-

schaft 2.
. (1934). Die pflanzlichen Plastiden als selbstaendige Elemente

der genetischen Konstitution. Bcr. Math. Phys. Kl. Saechs. Akad.

Wiss. 86.
. (1936). Zur Kenntnis der nichtmendelnden Buntheit der Laub-

blatter. Flora 30, 218-290.
Rensch, B. (1925a). Die Farbaberrationen der Voegel. Jour. Ornith.

Jahrg. 73.
. (19256). Untersuchungen zur Phylogenese der Schillerstruktur.

Jour. Oniith. Jahrg. 73.
Rhoades, Marcus M. (1931). Cytoplasmic inheritance of male sterility

in Zea Mays. Science 73.
. (1933). The cytoplasmic inheritance of male sterility in Zea

Mays. Jour. Genetics 27.
. (1936). The effect of varying gene dosage on aleurone colour in

maize. Jour. Genetics 33.

BIBLIOGRAPHY 351

Rhumbler, L. (1929). Zur Entwicklungsmechanik von Korkzieher-
geweihbildungen und verwandter Erscheinungen. Arch. Entwicklg-
mech. 119.

Richards, M. H., and E. Y. Furrow. (1922). A preliminary report on
the optic tract of eyeless flies. Univ. Oklahoma Bull. 2.

and . (1925). The eye and optic tract in Normal and Eye-
less Drosophila. Biol. Bull. 48.

Riedel, Hilda. (1935). Der Einfluss der Entwicklungstemperatur auf
Fliigel- und Tibialange von Drosophila melanogaster (wild, vestigial
und die reziproken Kreuzungen). Arch. Entwicklgmech. 132, 463-503.

Roberts, E. (1918). Fluctuations in a recessive Mendelian character and
selection. Jour. Exp. Zool. 27, 157-192.

Roesch, G. A. (1928). Experimented Untersuchungen iiber die Entste-
hung von Zwittern bei der Honigbiene (Apis mellifica L). Verhandl. d.
Deutsch. Zool. Ges.

Rokizky, P. Th. (1930). tiber die differentielle Wirkung des Gens auf
verschiedene Korpergegenden. Zeitschr. ind. Abstl. 67, 37-91.

Rueckert, J. (1899). Die erste Entwicklung des Eis der Elasmo-
branchier. Festschr. C. von Kupffer.

S

Saller, K., and F. Maroske. (1933). Chemische und genetische Unter-
suchungen an menschlichen Pigmenten, speziell demjenigen des Haares.

Zeitschr. Konstitutionsl. 17, 279-317.
Schaffner, J. H. (1934). Heredity and environment in relation to

character expression with special reference to intermittent characters

in Equisetum. Rev. Sud-Amer. Bot. 1, 8-17.
Schlottke, E. (1926). tiber die Variabilitaet der schwarzen Pigmen-

tierung und ihre Beeinflussbarkeit durch Temperaturen bei Habrobracon

juglandis Ashmed. Zeitschr. vgl. Physiol. 3.
Schmalfuss, H. (1928). Vererbung, Entwicklung und Chemie, nebst

entwicklungschemischen Untersuchungen an Organismen. Naturwiss.

16, 209-219.
and Helene Barth-Meyer, (Wilhelma Hinsch). (1931).

Vererbungstheoretische Betrachtungen nebst entwicklungschemischen

Untersuchungen iiber Vorkommen von Melanogen, insonderheit von

o-Dioxybenzol-Sfcoff, im Organismenreich. II. Zeitschr. ind. Abstl.

58, 332-371.
and H. Werner. (1926). Chemismus der Entstehung von Eigen-

schaften. Zeitschr. f. ind. Abstl. u. Vererb. 41.
Schmidt, M. (1931). Experimentelle Analyse der Genom- und Plasmon-

wirkung bei Moosen. II. Ueber eine hemihaploide und andere hetero-

ploide Rassen von Phy scorn itrium piriforme (L.) Brid. Ind. Abstl. -u.

Vererb. 57.
Schnakenbeck, W. (1921). Zur Analyse der Rassenmerkmale der

Axolotl. Arch. ind. Abstl. 27.
Schroeder, H. (1934). Untersuchungen ueber die Beeinflussung des

Bluetenfarbmusters von Petunia hybrida grandiflora hort. Jahrb.

Wiss. Bot. 79.

352 PHYSIOLOGICAL GENETICS

Schuwz, J. (1920). The minute reaction in the development of Droso-
phila melanogaster. (inn lies 14, 366 419.

— . (1935). Aspects of the relation between genes and development
in Drosophila. Amer. Naturalist 69.

. (1930). Variegation in Drosophila and the inert chromosome

regions. Proc. Nat. Acad. Sci. 22.
and T. DoBZHANSKY. (1933). Triploid hybrids between Droso-
phila melanogaster and Drosophila simulans. Jour. Exp. Zool. 65.
and . (1934). The relation of a dominant eye color in

Drosophila melanogaster to the associated chromosome rearrangement.

Genetics 19.
Schultz, W. (1915). Schwarzfaerbung weisser Haare. Arch. Entvricklg-

mech. 41.
. (1928). Kaltefarbung weisser Haare bei Schokolade-, Blau- und

Ganzweiss-Russenkaninchen. Kryptomerie und Manifestation der

Gene — Augenfiirbung — Explantatfarbung. Biol. Generalis 4, 291-320.
. (1929a). Kaeltepigmentierung von Albinohaar und -auge im

Reagensglase. Genetische Physiologic Pfluegers Arch. ges. Physiol.

221.
. (19296). Pigmentierung von Albinoaugen und sympathische

Heterochromie der Iris willkiirlich erzeugbar. Arch. f. Ophthalm.

122, 601-611.
. (1930). Kalteschwarzung weisser Haare und roter Albinoaugen,

ein Kapitel der genetischen Entwicklungsmechanik. Schr. Phys-

okon. Ges. Konigsberg. 67, 1-32.
. (1931). Methoden zur Darstellung versteckter mendelnder

Erbanlagen durch ihre Aktivierung ohne Kreuzung, besonders die

Kaltepigmentierung weisser Haare und roter Albinoaugen. Hdb. biol.

Arbeitsmeth. (Abderhalden) Abt. 9, Teil 3, 1055-1113.

(1932). Haarmelaninerzeugung bei Albinos innerhalb fiinf Minu-

ten unter dem Mikroskop und Weiteres zur Kalteschwarzung von Haar,

Haut und Auge. Arch. Dermatol. Syph. 165, 405-430.
Schwanwitsch, B. N. (1924). On the ground-plan of wing-pattern in

Nymphalids and certain other families of the Rhopalocerous Lepidop-

tera. Proc. Zool. Soc. London.
. (1925). On a remarkable dislocation of the components of the

wing-pattern in the Satyrid genus, Pierella. Entomologist 58.
. (1928a). Pierellisation of stripes in the wing-pattern of the

genus Rhaphicera Btl. (Lepidoptera, Satyridae). Zeitschr. f. Morph.

u. Oek. Tiere 11.
. (19286). Studies upon the wing-pattern of Pierella and related

genera of South American Satyridan butterflies. Zeitschr. f. Morph. u.

Oekol. 10.

(1929). Two schemes of the wing-pattern of butterflies. Zeitschr.

f. Morph. u. Oekol. Tiere 14.
Scott-Moncrieff, Rose. (1936). A biochemical survey of some

Mendelian factors for flower colour. Jour. Genetics 32.
Seidel, F. (1936). Entwicklungsphysiologie des Insekten-Keims. Ver-

handl. Deutsch. Zool. Ges. 38, Zool. Anz. Suppl. 9.

BIBLIOGRAPHY 353

Seiler, J. (1923). Geschlechtschromosomen-Unters. an Psychiden IV.

1923, die Parthenogenese der Psychiden. Biologische und zytologische

Beobachtungen. Zeitschr. ind. Abstl. 41.
. (1927). Ergebnisse aus der Kreuzung parthenogenetischer

und zweigeschlechtlicher Schmetterlinge. Biol. Zentralbl. 47.
Serebrovsky, A. S. (1929). A general scheme for the origin of mutations.

Amer. Naturalist 63, 374-378.
— . (1930). Untersuchungen iiber Treppenallelomorphismus. IV.

Transgenation scute-6 und ein Fall des "Nicht-Allelomorphismus"

von Gliedern einer Allelomorphenreihe bei Drosophila melanogaster.

Arch. Entwicklgmech. 122, 88-104.

and N. P. Dubinin. (1930). X-ray experiments with Drosophila.

Jour. Heredity 21.
Sexton, E. W., and C. F. A. Pantin. (1927). Inheritance in Gammarus

chevreuxi Sexton. Nature 119.
Seyster, E. W. (1919). Eye facet number as influenced by temperature

in the bar-eyed mutant of Drosophila melanogaster (ampelophila).

Biol. Bull. 37.
Shapiro, N. I. (1930). Step allelomorphism in Drosophila melanogaster.

VII. Jour. Exp. Biol. 6.
Shull, A. F. (1930a). Control of gamic and parthenogenetic reproduc-
tion in winged aphids by temperature and light. Zeitschr. ind. Abstl.

55, 108-126.
. (19306). Order of embryonic determination of the differential

features of gamic and parthenogenetic aphids. Zeitschr. ind. Abstl.

57, 92-111.
. (1935). Combinations of current and antecedent conditions in

relation to wing-production of aphids. Biol. Bull. 68, 35-50.
Sidorov, B. N. (1931). Investigations of the step-allelomorphism in

Drosophila melanogaster. Jour. Exp. Biol. 7.
Sinnott, E. W. (1921). The relation between body size and organ size

in plants. Amer. Naturalist 55.
. (1927). A factorial analysis of certain shape characters in squash

fruits. Amer. Naturalist 61.
. (1930). The morphogenetic relationships between cell and organ

in the petiole of Acer. Bull. Torrey Bol. CI. 57.
. (1935a). Evidence for the existence of genes controlling shape.

Genetics 20, 12-21.
. (19356). The genetic control of developmental relationships and

its bearing on the theory of gene action. Science 81, 420.
. (1936). A developmental analysis of inherited shape differences

in cucurbit fruits. Amer. Naturalist 70.
and L. C. Dunn. (1935). The effect of genes on the development

of size and form. Biol. Rev. 10, 123-151.
and Dorothy Hammond. (1930). Factorial balance in the deter-
mination of fruit shape in Cucurbita. Amer. Naturalist 64, 509-524.
, Helen Houghtaling, and A. F. Blakeslee. (1934). The

comparative anatomy of extra-chromosomal types in Datura stra-
monium. Carnegie Inst. Washington Publ. 451.

354 PHYSIOLOGICAL GENETICS

and S. Kaiskk. (1034). Two types of genetic control over the

development <>f shape. Hull. Torrey Bat. CI. 61, 1-7.
Sikks, M. J. (1031a). Plasmatic influences upon the inheritance in Vicia

Faba. 1. The elimination <>f a whole Linkage-group in the plasm of

Vicia Faba Minor. Kon. Ak. Wet. Amsterdam Proc. 34, 1057-1062.
. (19316). Plasmatic influences upon the inheritance in Vicia Faba.

II. Different plasmatic reactions upon identical genotypes. Kon. Ak.

Wet. Amsterdam /'roc. 34, 1164-1 17J.

(1931c). Plasmatic influences upon the inheritance in Vicia Faba.

III. The elimination of a definite factor (variegated) as caused by the
type of plasm. Kon. Ak. Wet. Amsterdam Proc. 34, 1340-1346.

Sivertzev-Dobzhansky, N. P., and Tn. Dobzhansky. (1933). Deficiency
and duplications for the gene bobbed in Drosophila melanogaster.
Genetics 18.

Skalinski, M. (1924). Recherches sur la selection des bourgeons chez
une race polymorphe du Petunia et le probleme de la production de
nouvelles vari6tes (Polish, French summary). Mem. Inst. Gen.
Varsovie 2.

. (1928). Sur les causes d'une disjonction non typique des hybrides

du genre Aquilegia. Acta Soc. Bot. Polon. 6, 141-173.

Smith, E., and E. C. MacDowell. (1931). The differential effect of
hereditary mouse dwarfism on the anterior-pituitary hormones. Anat.
Rec. 60.

Snell, G. D. (1931). Inheritance in the house-mouse, the linkage rela-
tions of short-ear, hairless, and naked. Genetics 16.

, E. Bodemann, and W. Hollander. A translocation in the house

mouse and its effect on development. Jour. Exp. Zool. 67.

and D. I. Pic-ken. (1935). Abnormal development in the mouse

caused by chromosome unbalance. Jour. Genetics 31.
Speicher, B. R. (1932). Dominance of two kidney allelomorphs in

Habrobracon juglandis (Ash.). Biol. Bull. 63, 372-380.
. (1933). A morphological study of the effective period of "eyeless"

and "glass" in Habrobracon. Univ. Pittsburgh Bull. 30, No. 2.

(1934). The temperature-effective period in the development of

"eyeless" in Habrobracon. Amer. Naturalist 68. (Abstract.)
Spencer, Warren P. (1930). Mosaic orange — an asymmetrical eye

color in Drosophila Hydei. Jour. Exp. Zool. 56.
Spoettel, W. (1914). tiber die Farben der Vogelfedern. II. Die

Faerbung der Columba livia etc. Zool. Jahrb. (Anat.). 38.
Stadler, L. J. (1932). On the genetic nature of induced mutations in

plants. Proc. 6th Internal. Congr. Genet. 1.
Standfuss, M. (1896). Handbuch der palaearktischen Grossschmetter-

linge. Jena.
. (1907). Juengste Ergebnisse aus der Kreuzung verschiedener

Arten und der Paarung zweier Lokalrassen derselben Art. Mitt. d.

Schweizer. Entomol. Ges. 4.
. (1908). Experimentelle zoologische Studien. Denkschr. Schwei-

zer Ges. Naturvriss.

BIBLIOGRAPHY 355

. (1914). Mitteilungen zur Vererbungsfrage. Mitt. Schweizer

Entomol. Ges. Bd. 12.
Stanley, W. F. (1931). The effect of temperature on vestigial wing in

Drosophila melanogaster with temperature-effective periods. Physiol.

Zool. 4.
. (1935). The effect of temperature upon wing size in Drosophila

melanogaster. Jour. Exp. Zool. 69.
Stein, Emmy. (1932). Uber den durch Radiumbestrahlung von Embry-

onen erzeugten erblichen Krankheitskomplex der Phytocarcinome von

Antirrhinum niajus. Phytopath. A, 4.
. (1935o). Uber eine Spontan-Mutante von Antirrhinum majus:

Mut. pseudochimera und die Entartung ihrer Embryonalgewebe.

Zeitschr. ind. Abstl. 69.

(19356). Weitere Analyse der Gruppe A von den durch Radium-

bestrahlung veraenderten Erbanlagen bei Antirrhinum. Kern- und

Zellveraenderung in der krebsigen Gewebeentartung. Zeitschr. ind.

Abstl. 69.
Steiner, Hans. (1932). Vererbungsstudien am Wellensittich Melo-

psittacus undulatus (Shaw). Arch. Ind. Klaus Stift. 7.
Stella, E. (1934). Figura ed istogenesi delle penne, influenze ormonalie

genetiche nei polli di razza "Golden laced Wyandotte." Arch. Zool.

Ital. 20, 237-264.
Stern, C. (1926). Eine Kreuzungsanalyse von Koerperfaerbungen von

Drosophila melanogaster, verbunden mit drei neuen Allelomorphen des

Faktors "ebenholz." Zeitschr. ind. Abstl. 41.
. (1927). Der Einfluss der Temperatur auf die Ausbildung einer

Fluegelmutation bei Drosophila melanogaster. Biol. Zentralbl. 47.
. (1929a). Kleinere Beitraege zur Genetik von Drosophila melano-
gaster. I. Ein Hemmungsfaktor der Purpuraugenfarbe. Zeitschr.

ind. Abstl. 62.
. (19296). Uber die additive Wirkung multipler Allele. Biol.

Zentralbl. 49, 261-290.

. (1930). Multiple Allelie. Handbch. d. Uererbgswiss.

. (1935). The behavior of unstable genie loci — an hypothesis.

Proc. Nat. Acad. Sci. 21.
. (1936). Somatic crossing over and segregation in Drosophila

melanogaster. Genetics. 21.
and V. Rentschler. (1936). The effect of temperature on the

frequency of somatic crossing over in Drosophila melanogaster. Proc.

Nat. Acad. Sci. 22.
Stockard, C. R. (1910). The influence of alcohol and other anesthetics

on embryonic development. Amer. Jour. Anal. 10.
. (1913). The artificial production of structural arrest etc. Proc.

N. Y. Pathol. Soc. 13.
. (1921). Developmental rate and structural expression. Amer.

Jour. Anat. 28.
. (1921-1922). The significance of modifications in body structure.

Harvey Soc. Lect.

356 I'/IYSIOLOGICAL GENETICS

— . (1034). Internal constitution and genk factors in growth

determination. Cold Spring Harbor Symp. on Quant. Biol. II, 118-127.
Stroeek, F. II. (1936). The gene, its function and its meaning in genetics.
Quart, lur. Biol. 11.

Stkohl, J., and W. KoBHLEB. (1934). Experimentelle Untersuchungen

uebei die Entwicklungsphysiologie tier Fluegelzeichnung bei der

Mehlmotte. Verhandl. Schiveizer. Naturforsch. Ges. Zurich.
Sturtev.wt, A. H. (1920). The vermilion gene and gynandromorphism.

Proc. Soc. Exp. Biol. Mi<{. 17.

. inheritance of direction of coiling in Lymnaea. Science 68.

. (1925). The effects of unequal crossing over at the bar locus in

Drosophila. Genetics 10.
. (1927). The effect of the Bar-gene of Drosophila in mosaic eyes.

Jour. Exp. Zool. 46.
. (1928). A further study of the so-called mutation at the bar locus

of Drosophila. Genetics 13, 401-409.
. (1929). The claret mutant type of Drosophila simulans: a study

of chromosome elimination and of cell-lineage. Zeitschr. Wiss. Zool.

135, 323-356.
. (1932). The use of mosaics in the study of the developmental

effects of genes. Proc. Qth Internal. Congr. Genetics 1, 304-307.
and T. H. Morgan. (1923). Reverse mutation of the bar gene

correlated with crossing over. Science, N.S. 67.

and J. Schultz. (1931). The inadequacy of the sub-gene hypothe-

sis of the nature of the scute allelomorphs of Drosophila. Proc. Nat.
Acad. Set. 17, 265-270.
Sueffert, F. (1925). Geheime Gesetzmaessigkeiten in der Zeichnung
der Schmetterlinge. Rev. Suisse Zool. 32.

■■ — . (1927). Zur verleichen den Analyse der Schmetterlingszeiehnung.

Biol. Zentralbl. 47.

. (1929a). Die Ausbildung des imaginalen Fluegelschnittes in der

Schmetterlingspuppe. Zeitschr. f. Morph. u. Oekol. 14.

(19296). Morphologische Erscheinungsgruppen in der Fluegel-

zeichnung der Schmetterlinge, insbesondere die Querbindenzeichnung.
Arch. Entwicklgmech. 120.
Surrarrer, T. C. (1935). The effect of temperature on a mottled-eye
stock of Drosophila melanogaster. Genetics 20.

Tammes, Tine. (1934). Die Ubereinstimmung zwischen Entwicklungssta-
dien und Phiinotypen verschiedener Genotypen. Genetica 15, 519-528.

Tanaka, Y. (1913). A study of Mendelian factors in the silkworm,
Bombyx mori. Jour. Coll. Agri. Tohoku Imp. Univ. Sapporo 5.

. (1916). Genetic studies on the silkworm. Jour. Coll. Agric.

Tohoku Imp. Univ. Sapporo 7.

(1924). Maternal inheritance in Bombyx mori. Genetics Bd. 9.

Tenenbaum, E. (1933). Zur Vererbung des Zeichnungsmusters von
Epilachna chrysomelina F. Biol. Zentralbl. 53.

BIBLIOGRAPHY 357

Terao, H. (1917). On reversible transformability of allelomorphs.

Am. Naturalist 51.
Thompson, D'Arcy. (1916). Growth and form. London.
Thompson, D. H. (1931). The side-chain theory of the structure of the

gene. Genetics 16.
Timofeeff-Ressovsky, Helene. (1931). tjber phanotypische Manifes-

tierung der polytropen (pleiotropen) Genovariation Polyphaen von

Drosophila funebris. Naturwiss. 19, 765-768.
and N. W. Timofeeff-Ressovsky. (1926). tlber das phaeno-

typische Manifestieren des Genotyps. II. Lvber idio-somatische

Variations-gruppen bei Drosophila funebris. Arch. Entwicklgmech. 108.
Timofeeff-Ressovsky, N. W. (1925). On the phenotypic realization of

the genotype. I. The geno-variation radius incompletus in Drosophila

funebris. (Russian.) Jour. Exp. Biol. 1.
. (1927). Studies on the phenotypic manifestation of hereditary

factors. I. On the phenotypic manifestation of the genovariation

radius incompletus in Drosophila funebris. Genetics 12.
. (1928). The effect of temperature on the formation of the cross-
veins on the wings of a genovariation in Drosophila funebris. (Russian.)

Jour. Exp. Biol. 4.
. (1929a). Der Einfluss der Temperatur auf die Ausbildung der

Queradern an den Fliigeln bei einer Genovariation von Drosophila

funebris. Jour. Psych. Neur. 38, 134-1-46.
. (1929b). The phenotypic realization of the gene vti in Drosophila

funebris. Proc. Allruss. Cong. Gen. 2.
. (1931). Gerichtetes Variieren in der phanotypischen Mani-

festierung einiger Genovariationen von Drosophila funebris. Naturwiss.

19, 493-497.
. (1932). Verschiedenheit der "normalen" Allele der white-Serie

aus zwei geographisch getrennten Populationen von Drosophila melano-

gaster. Biol. Zentralbl. 62, 468-476.
. (1933a). Mutations of the gene in different directions. Proc. 6th

Internal. Cong. Gen. 1.
. (19336). Rueckmutationen und die Genmutabilitaet in ver-

schiedenen Richtungen. III. Roentgenmutationen in entgegenge-
setzten Richtungen am forked-Locus von Drosophila melanogaster.
Zeitschr. ind. Abst. Vererb. 64.

- — . (1933c). Rueckmutationen und die Genmutabilitaet in ver-
schiedenen Richtungen. IV. Roentgenmutationen in verschiedenen
Richtungen am white-Locus von Drosophila melanogaster. Zeitschr.
ind. Abst. Vererb. 65.

— . (1933d). Riickgenmutationen und die Genmutabilitat in ver-
schiedenen Richtungen. V. Gibt es ein wiederholtes Auftreten identi-
scher Allele innerhalb der white-AUelenreihe von Drosophila melano-
gaster? Zeitschr. ind. Abstl. 66, 165-179.

— . (1934). tJber den Einfluss des genotypischen Milieus und der
Aussenbedingungen auf die Realisation des Genotyps. Genmutation
vti (venae transversae imcompletae) bei Drosophila funebris. Nachr.

358 PHYSIOLOGICAL GENETICS

Ges. Wiss. Gdttingen Math.-Phyaik. Kl. Fachgr. VI N.F. 1, Nr. G,
53- 106.
. (1935). tJber"muetterliche Vererbung"bei Droeophila. Natur-

iriss. 23.

ami H. A. Timui i.i.i i -Kkssovsky. (1934). Polare Schwankungen
in der phaenotypischcn Manifestierung einiger Genmutationen bei
Drosophila. Zeitschr. ind. Abst. Vererb. 67.

, K. CJ. Zimmek, and M. Delbruck. (1935). Uber die Natur der

Genmutation und der Genstruktur. Nachr. Ges. Wiss. Gdttingen
Math.-Phys. Kl. 6, Biol. 1, L89-245.
Toblek, Margarets. (1931). Experimentelle Analyse der Genom- und
Plasmonwirkung bei Moosen. IV. Zur Variabilitat des Zellvolumens

einer Sippenkreuzung von Funaria hygrometrica und deren bivalenten
Rassen. Zeitschr. ind. Abstl. 60, 39-62.

Toldt, K. (1912). Beitraege zur Kenntnis der Behaarung der Saugetiere.
Zool. Jahrb. (Anat.) 33.

Tower, W. L. (1906). An investigation of evolution in chrysomelid
beetles of the genus Leptinotarsa. Carnegie Inst. Publ. Washington 48.

. (1910). The determination of dominance and the modification of

behavior in alternative (Mendelian) inheritance, by conditions sur-
rounding or incident upon the germ-cells at fertilization. Biol. Bull. 18.

Toyama, K. (1906a). Mendel's laws of heredity as applied to the silkworm
crosses. Biol. Zentralbl. 26.

. (19066). Studies on the Hybridology of insects: I. On some

silkworm crosses, with special reference to Mendel's laws of heredity.
Bull. Coll. Agric. Tokyo Univ. 7.

. (1909). A sport of the silkworm and its heredity behaviour.

Zeitschr. ind. Abst. Vererb. 1.

Troland, L. T. (1917). Biological enigmas and the theory of enzyme
action. Amer. Naturalist 51.

Twitty, V. C. (1934). Growth correlations in amphibia studied by the
method of transplantation. Cold Spring Harbor Symposia. 2.

. (1936). Correlated genetic and embryological experiments on

Triturus. I. Hybridization: Development of three species of Triturus
and their hybrid combinations. II. Transplantation: The embryo-
logical basis of species differences in pigment pattern. Jour Exp.
Zool. 74.

and H. A. Elliott. (1934). The relative growth of the amphibian

eye, studied by means of transplantation. Jour. Exp. Zool. 68.

U

Ubisch, L. v. (1936). tlber die Organisation des Seeigelkeims. Arch.
Entwicklgmech . 134.

Van Assen, J. (1930). Angeborene Kyphose. Acta Chir. Scandinavica
67.

BIBLIOGRAPHY 359

W

Waardenburg, P. J. (1930). Die Zurueckfuehrung einer Reihe erblich-

angeborener familiaerer Augenmissbildungen auf eine Fixation normaler

fetaler Verhaeltnisse. Graefes Arch. 124.
Wakakuwa, Sh. (1934). Embryological studios on the different seed-
development in reciprocal interspecific crosses of wheat. Japanese

Jour. Bot. 7, 151-184.
Warren, D. C. (1924). Inheritance of egg size in Drosophila melano-

gaster. Genetics 9.
. (1933). Retarded feathering in the fowl. Jour. Heredity 24,

431-434.
Weinstein, A. (1928). Heredity and development. Chemistry in

Medicine, New York.
Weiss, Paul. (1935). The so-called organizer and the problem of

organization in amphibian development. Physiol. Rev. 15.
Werber, E. J. (1916). Experimental studies on the origin of monsters.

Jour. Exp. Zool. 21.
Wettstein, F. von. (1924a). Morphologie und Physiologie des Form-

wechsels der Moose auf genetischer Grundlage I. Zeitschr. f. ind.

Abstammungs-u. Vererb. Bd. 33.
. (19246). Kreuzungsversuche mit multiploiden Moosrassen II.

Biol. Zentralbl. Bd. 44.
. (1928a). Morphologie und Physiologie des Formwechsels der

Moose auf genetischer Grundlage. II. Bibliotheca Genetica 10,

1-216.
. (19286). liber plasmatische Vererbung und iiber das Zusam-

menwirken von Genen und Plasma. Ber. Deutsch. Bot. Ges. 46,

32-49.
. (1930). tJber plasmatische Vererbung, sowie Plasma und Gen-

wirkung, II. Nachr. Ges. Wiss. Gottingen Math.-Phys. Kl. Fachgr.

VI, Nr. 2.

(1934). Uber plasmatische Vererbung und das Zusammenwirken

von Genen und Plasma. Wiss. Woche Frankfurt ,M. 2.-9.9.1934 1,

Erbbiologie. 31-36.
Whiting, A. R. (1934). Eye colours in the parasitic wasp Habrobracon

and their behavior in multiple recessives and in mosaics. Jour. Genetics

29.
Whiting, P. W., R. J. Greb, and B. R. Speicher. (1934). A new type

of sex-intergrade. Biol. Bull. 66.
Wigglesworth, V. B. (1934). The physiology of ecdysis in Rhodnius

prolixus (Hemiptera). II. Factors controlling moulting and "meta-
morphosis." Quart. Journ. Microsc. Sc. 77.

. (1935). Functions of the corpus allatum of insects. Nature.

Winkler, H. (1923). Uber die Rolle von Kern und Protoplasma bei der

Vererbung. Verh. deutsch. Ges. Vererb. Muenchen.
. (1934). Uber zwei Solanum-Chimaren mit Burdonen-epidermis.

Planta 21, 613-656.

360 PHYSIOLOGICAL GENETICS

de Winton, D., and J. B. S. Haldane. (1933). The genetics of Primula

sinensis. II. Segregation and interaction <»t" factors in the diploid.

Jour. Genetics 27, l 1 1
Wnsciu, K. (1929). Studies on sex-differentiation anil sex-determination

in amphibians. 1. II. III. Jour. Exp. Zool. 52, 64.
Wolsky, A., and J. S. Huxley. (1934). The structure and development

of normal and mutant eyes in Gammarus chevreuxi. Proc. Roy. Soc. B

114, 3<>4 392.

and . (1936). The structure of the non-facetted region In

the bar-eye mutants of Drosophila, and its bearing on the analysis of
genie action upon Arthropodan eyes. Proc. Zool. Soc. London.

Wolsky, A. (1936). Uber Zusammenhaenge zwischen Entwicklungs-

physiologie und Genetik in der experimentellen Morphologic Arb.

Ung. Biol. Forschgsinst. 8.
Wriedt, Ch., and O. L. Mohr. (1928). Amputated, a recessive lethal

in cattle. Jour. Genetics 19.
Wright, S. (1916). An intensive study of the inheritance of color etc.

Carnegie Inst. Washington Publ. 241.
. (1917-1918). Color inheritance in mammals. Jour. Heredity Bd.

8, 9.
. (1925). The factors of the albino-series of guinea pigs etc.

Genetics Bd. 10.
. (1927). The effects in combination of the major color-factors of

the guinea pig. Genetics 12, 530-569.

. (1929a). Fisher's theory of dominance. Amer. Naturalist 63.

. (19296). The evolution of dominance. Amer. Naturalist 63.

. (1929c). The dominance of bar over infra-bar in Drosophila.

Amer. Naturalist 63, 479-480.
. (1930). The genetical theory of natural selection. (A review.)

Jour. Heredity 21.
. (1932). Complementary factors for eye color in Drosophila.

Amer. Naturalist 66, 282-283.
. (1934a). An analysis of variability in number of digits in an

inbred strain of guinea pigs. Genetics 19, 506-536.
. (19346). Genetics of abnormal growth in the guinea pig. Cold

Spring Harbor Symposia on Quantitative Biology 2, 137-147.
. (1934c). On the genetics of subnormal development of the head

(otocephaly) in the guinea pig. Genetics 19, 471-505.
. (1934a7). Physiological and evolutionary theories of dominance.

Amer. Naturalist 68, 24-53.
. (1935). A mutation of the guinea pig, tending to restore the

pentadactyl foot when heterozygous, producing a monstrosity when

homozygous. Genetics 20, 84-107.

and O. N. Eaton. (1929). The persistence of differentiation

among inbred families of guinea pigs. U. S. Dept. Agric. Washington
Tech. Bull. 103.

, H. C. McPhee, and O. N. Eaton. (1932). Hereditary variations

of the guinea pig. Proc. 6th Internal. Congr. Genet. 2.

BIBLIOGRAPHY 361

and K. Wagner. (1934). Types of subnormal development of

the head from inbred strains of guinea pigs and their bearing on the

classification and interpretation of vertebrate monsters. Amer. Jour.

Anat. 54, 383-447.
Wrinch, D. M. (1934). Chromosome behavior in terms of protein

pattern. Nature 134.
. (1936). On the molecular structure of chromosomes. Proto-

plasma 25.
Wulkopf, H. (1936). Hitze- und Frostreize in ihrer Wirkung auf das

Fluegelmuster der Mehlmotte Ephestia kuehniella Z. Arch. Entwicklg-

mech. 134.

Zeleny, C. (1918). Germinal changes in the bar-eyed race of Drosophila

during the course of selection for facet number. Proc. Indiana Acad.

Set.
. (1919). A change in the bar gene of Drosophila involving further

decrease in facet number and increase in dominance. Jour. Gen.

Physiol. 2.
. (1920). A change in the bar gene of Drosophila melanogaster

involving further decrease in facet number and increase in dominance.

Jour. Exp. Zool. 30.
. (1923). The temperature coefficient of a heterozygote etc.

Biol. Bull. Marine Biol. Lab. Bd. 44.
— . (1933). Genetics and embryology. Science 77.

Zimmermann, W. (1934). Genetische Untersuchungen an Pulsatilla
I-III. Flora N.F. 29.

AUTHOR INDEX

Abbe, 38
Abderhalden, 99
Agol, 165, 236, 296
Anderson, 38, 80, 97, 187, 295
Anderson-Kottoe, 291
Artom, 153

Astauroff, 70, 208, 225
Auerbach, 28

B

Bagg, 24, 47, 49, 50

Bailey, 188

Balkashina, 36, 37

Baltzer, 44, 180, 181, 194, 268

Bateson, 99, 123

Baur, 156, 176, 273, 283, 292

Beadle, 133, 164, 165, 166, 185, 280,

306
Becker, 256, 272
Beebe, 5
Beliajeff, 225
Berg, 284

Bergraann, 310, 314, 315
Berndt, 260
Blakeslee, 151, 213
Blyth, 190
Bobroff, 9, 40
Bodenstein, 189
Boeke, 41

Bonnevie, 24, 25, 51, 222
Bonnier, 182
Boveri, 206, 267, 280
Boycott, 206, 224
Braun, 243, 244
Breitenbecher, 226
Bridges, 36, 44, 107, 125, 133, 140,

148, 149, 151, 152, 168, 174, 175,

207, 208, 225, 284

Brieger, 280
Brink, 94, 280, 303
Buddenbrock, 38, 221
Bytinsky-Salz, 189

Callow, 188
Caspari, 183
Castle, 54, 56, 161, 193, 208, 213,

258, 300
Charles, 161
Chen, 30, 31, 76, 216
Chesley, 24
Child, C. M., 19, 46, 48, 50, 51, 72,

203, 236, 238, 239, 240, 296
Child, G., 237
Chittenden, 279
Clancy, 187
Clausen, 268
Cockayne, 130
Conklin, 206
Correns, 99, 149, 269, 278, 283, 288,

289, 290, 291, 292, 295
Coyne, 164
Crampton, 206
Crew, 66, 107, 108, 164, 165
Csik, 161, 191, 212
Cuenot, 91, 108, 160

I)

Dahlberg, 226

Danforth. 10, 26, 46, 219, 226

Darby, 227

Davenport, 99, 265

David, 25

Dawes, 227

Delbrueck, 293, 294, 316

Dellingshausen, 275

De Meijere, 130, 234, 251

363

:W4

PIIYSlOUHilCAL GENETICS

Demerec, 17 1. 176. l'.u. 283, 288,

289, 290, 29"), 299
De Roche, L80
De Vries, 63, 288
De \\ mt. hi. 80, 187
Diver, 224
Dobrovolakaja, 24
Dobzhansky, 38, 66, 68, 109. 114,

183, 207
Doerries-Rueger, 272
Doncaster, 250
Dorfmeister, 9
Driesch, 201

Driver, 11, 16, 17, 75, 139, 216
Dry, 220
Dubinin, 236, 237, 239, 240, 296,

305, 306, 307, 308
Duncan, 38, 207
Duncker, 36, 222
Dunn, 10, 25, 26, 46, 107, 161, 164,

214, 215, 258
Du Toit, 10, 26

E

East, 268, 274, 276

Eaton, 49

Eker, 81

Elliott, 221

Emerson, 176

Emsweller, 280

Engelhardt, 197, 198

Engelsmeier, 155

Ephrussi, 44, 133, 164, 165, 166, 181,

185, 186, 187
Euler, 93
Eyster, 288, 295

Faulkner, 218, 259

Federley, 124, 233, 235, 250

Feldman, 210

Feldotto, 247

Ferwerda, 113, 173

Fischer, Eugen, 78

Fischer, E. V., 6, 9, 108, 130, 250

Fisher, 108, 121, 122, 123, 131

Ford, 46, 53, 130, 251

Fraps, 148, 217, 218, 219, 259

Fraser, 260

Friesen, 9, 16, 45, 106, 237, 239, 240,

302, 303
Frost, 280
Fryer, 130, 234, 251
Fukui, 260

G

Gabritschevsky, 36, 207

Gay, 290

Geigy, 9

Gerhardt, 149, 256

Gerould, 130, 250

Gerschenson, 296

Giersberg, 196

Glass, 297, 307

Gley, 182

Godlevski, 267

Goernitz, 34

Goodrich, 23, 44, 53, 54, 259, 260

Goodspeed, 268

Gordon, 260

Gortner, 91

Gottschewski, 9

Gowen, 149, 280, 284, 290

Graubard, 91, 92

Greb, 177

Green, 108

Greenwood, 190

Gregory, 54, 193, 208, 213

Gross, 153

Grueneberg, 306

Gurwitsch, 204

Gustafson, 218, 259

Guthrie, 225

Guyenot, 204

Guyer, 148, 283

H

Haberlandt, 177
Hadjidimitroff, 159
Hadorn, 181, 268
Haecker, 23, 34, 226, 257
Haemmerling, 179, 182, 192, 200,
265, 277

AUTHOR INDEX

365

Hagedoorn, 282, 298

Hagiwara, 97

Haldane, 66, 93, 98, 121, 123, 130,

188, 288, 292
Hallquist, 93
Hamburger, 180
Hammett, 193
Hammond, 217
Hansen, 53, 54, 260, 293
Harder, 268
Harland, 108
Harnly, 12, 18, 30, 42, 46, 54, 103,

128
Harrison, R. G., 47, 200, 205, 220
Harrison, T. W. H., 250
Hashimoto, 53
Henke, 10, 20, 21, 23, 196, 197, 199,

210, 231, 233, 234, 245, 246, 247,

248, 250, 251, 252, 253, 257, 258,

261, 262
Hersh, 11, 75, 103, 104, 106, 114,

128, 139, 143, 144, 167, 210, 211,

215, 216
Heys, 293
Hinrichs, 47
Hirata, 261
Hoener, 110, 118
Hoerstadius, 267
Hoge, 70
Holmes, 259
Holtfreter, 47, 179, 190
Holthausen, 313
Honing, 46, 112, 113, 273
Houghtaling, 212, 213, 218
Huxley, 32, 34, 46, 53, 66, 177, 182,

196, 209, 210, 211, 215, 216

Jones, 280

Juhn, 217, 218, 219, 259

Just, 265

K

Kaiser, 218

KamenofT, 215

Karczag, 283, 292

Karrer, 93

Katsuki, 10, 280

Kellicott, 47

Kellogg, 99

Kikkawa, 108

King, 225

Kniesche, 34

Kobozieff, 24

Koehler, 243, 244, 247

Kofoid, 206

Koller, 91

Koltzoff, 264, 265, 282, 293, 300,
301, 316

Komai, 70

Kosminsky, 40, 194, 233

Kosswig, 44, 86

Kostoff, 268, 292

Krafka, 11, 20, 30, 74, 75

Krieg, 257

Krueger, 60, 61, 62

Kuehn, 10, 21, 23, 45, 78, 127, 183,
184, 185, 189, 196, 197, 198, 199,
210, 231, 233, 234, 235, 243, 246,
247, 250, 251, 252, 253, 258, 273

Kuester, 256

Ibsen, 258
Iljin, 155, 258
Imai, 86, 176, 288
Ives, 20

Johannsen, 30, 32, 156
Jollos, 9, 277, 278

Lammertz, 268
Lamy, 107, 108, 164, 165
Lancefield, 86, 107
Landauer, 10, 25, 107, 214
Lang, 112, 224
Lawrence, 97, 150
League, 137
Lebedeff, 107, 189
Lehmann, 50, 51, 208, 274
Lesley, 280
Lewis, 47

306

PUYSIOUXilCAL CEN ETICS

Lewit, 230, 2%
Lilienfeld, 291
Lillie, 204, 218, 219
Lindahl, 205
Linden, 45

Lindstrom, 149, 212, 218
Little, 24, 47, 49, 50, 91, 271
Luce, 11, 106, 140, 143
Ludwig, 221, 224, 225, 227
Lynch, 207

M

Newman, 49, 222, 225, 226
Nichols, 260
Nileon-Ehle, 93, 150

Nujdin, 239

O

Ochlkcrs, 63, 64, 65, 69
Offermann, 137, 305, 309

Ogura, 77, 118, 127
Onslow, 91, 93, 97

McClendon, 47

McDowell, 188

Machida, 250

Mangelsdorf, 148, 280

Margolis, 76, 138, 216

Mark, 293, 301

Martini, 212

Medvedev, 26, 30, 31, 187

Melchers, 272

Merrifield, 9

Meyer, 293, 301

Michaelis, 274, 275, 277, 278

Milner, 97

Minami, 130, 195, 231

Mittasch, 282

Mohr, 41, 49, 56, 57, 58, 62, 66, 80,

87, 109, 117, 118, 123, 133, 134,

137
Montalenti, 49, 219
Morgan, L. V., 137, 138
Morgan, T. H., 70, 108, 174, 188,

206, 207, 225, 284
Muentzing, 55, 152, 153
Muller, 114, 122, 123, 130, 131, 134,

135, 137, 140, 282, 284, 286, 287,

293, 297, 299, 306, 307, 309, 316
Murray, 271

X

Nadler, 12
Nagai, 77
Nagamori, 77
Nawashin, 149, 268

Pantin, 206

Pariser, 125

Parkes, 188

Patterson, 174, 175, 306

Pearsall, 216

Pearson, 222

Pellew, 279

Pernitsch, 23

Philipchenko, 38

Phillips, 258

Plagens, 24

Plagge, 183, 184, 185

Plough, 313

Plunkett, 19, 46, 66, 74, 75, 122, 190,

199, 225, 236, 238, 282
Pol, 41

Poppelbaum, 235
Poult on, 234, 251
Prokofieva, 284, 306, 309
Przibram, 226, 227
Punnett, 188

R

Raffel, 306
Rasmusson, 150
Redfield, 207, 208
Renner, 300
Rensch, 35
Rhoades, 149
Rhumbler, 77
Riedel, 12, 82
Roberts, 12

AUTHOR INDEX

367

Robertson, 76
Robinson, 93, 97
Roesch, 10
Rueckert, 264

S

Sando, 97

Schlottke, 233

Schmalfuss, 92, 113, 282, 286, 299,

309
Schnakenbek, 23
Schotte, 179
Schultz, J., 32, 33, 91, 149, 162, 163,

164, 165, 236, 237, 239, 240, 296,

307
Schultz, W., 155, 161
Schwanwitsch, 196, 246, 248
Schwemmle, 274
Scott-Moncrieff, 94, 96, 97, 98, 114,

120, 150, 169
Seidel, 178
Seiler, 153
Serebrowskaya, 284
Serebrowsky, 53, 236, 295, 296, 298,

299, 305
Sexton, 206
Seyster, 11
Shapiro, 284
Sherman, 97
Shriner, 97
Sidorov, 306

Sinnott, 46, 213, 215, 216, 217, 218
Sirks, 272, 273
Skalinska, 273, 279
Smith, 188
Snell, 108
Speicher, 26, 177
Spemann, 47, 179, 195, 268
Spoettel, 34
Stadler, 298, 299
Standfuss, 4, 5, 9, 124, 128, 233, 234,

250
Stanley, 12, 13, 17, 18, 109
Staudinger, 293
Stein, 25
Steiner, 35, 36

Stern, 84, 85, 89, 114, 118, 120, 132,
134, 136, 137, 145, 150, 176, 264,
287, 292

Stockard, 10, 47, 48, 50

Stone, 306

Sturtevant, 108, 109, 137, 138, 140,
143, 174, 175, 177, 183, 206, 207,
236, 237, 239, 240, 296, 304, 306,
307, 309

Sueffert, 196, 233, 234, 243, 246, 248

Tammes, 40, 46
Tanaka, 77, 206, 207, 251, 280
Tchetverikoff, 107, 156
Tenenbaum, 247, 251
Thompson, D'Arcy, 209, 211
Thompson, 297
Timofeeff-Ressowsky, 5, 66, 69, 71,

109, 123, 207, 223, 230, 286,

293, 316
Toldt, 257
Tower, 5, 99, 251
Toyama, 184, 206, 207, 251, 277
Troland, 282
Twitty, 179, 220, 221, 259

Van Assen, 25
Van Cleave, 212
Van Overbeek, 62
Volotov, 305
Von Ubisch, 205, 267

W

Waardenburg, 41
Waddington, 202
Wagner, 49
Ward, 103
Warren, 53, 207
Weiss, 204
Werber, 47
Werner, 92
Wertz, 274
West, 61

368

I'llYSlolJHiHM. (1ENETICS

Wettstein, 56, L53, L54, 271,

276, 277, 278, 286
Wheldale, 93, 97
Whiting, 164, 177, 183
Wiggle8worth, 38, 77, 221
Wilder, 222
Willstaetter, 93
Wilson, 206
Winge, 261
Winkler, 275
Wirtz, 278
Witschi, 181
Wolsky, 32, 34, 76, 177
Wriedt, 41, 49

272, Wright, 46, 49, 50, 52, 66, 70, 71, 72,
73, 81, 87, 88, 89, 91, 100, 101,
102, los, 114,118, 121, 123, 143,
157, 158, 159, 100, Kil, 104, 165,
170, 220, 258, 282, 287, 297

Wrinch, 301, 302, 310, 314, 316

Wulkopf, 247

Zeleny, 11, 16, 45, 46, 73, 75, 104,

Hi:., 114, 137, 143, 297
Zimmer, 293, 310
Zimniormann, 10, 200

SUBJECT INDEX

Abraxas grossulariata, 22, 198, 247
Abraxas grossulariata lacticolor, 250
Acanthocephala, 212
Acetabidaria, 182, 192, 200
Activators, 182
Allot riploid hybrids, 149
Allium fistulosum, 280
Amphibia, merogony, 44

ovocytes, 264

species hybrids, 180
Anemone, 200
Anthocyanin, 169-170
Antimorphs, 135
Antirrhinum, A. majus, 98

geographic species, 273

radium-induced abnormalities, 25
Aquilegia vulgaris, 38, 279
Argynnis, paphia, 130

paphia-valesina, 250
Arrhenius's formula, 44
Artemia salina, 153
Ascaris, chromatin diminution, 280
Auxin, 62
Axolotl, development, 23

B

Bacteriophage, 60, 61, 74
Balance of the genes, 133, 148
Bantam cocks, hen-feathered, 188
Barley, chemistry of chlorophyll-
defective mutants, 93
Baroid translocation, 305
Basidiomycetes, 268
Birds, development of abnormalities,
24-26
feather pigmentation, 34-36
See also Feather
Blood-groups, 99

Bombyx (see silkworm)
Brachyphalangy, 41

Bruchus, 226

Butterflies, experiments on, 45, 47
morphology of pattern, 246
pattern and form of scales, 234
seasonal polymorphism, 249
wing pattern, 4

Callimorpha dominula, 250

Canna, 273

Capsicum, 218

Carassius auratus, 260

Caterpillar with pupal antennae, 173

Cats, with one blue and one yellow

eye, 226
Cattle monstrosities, 49
Cell-constant animals, 212
Chain molecules, 293, 301-302
Chemodifferentiation, 182, 265
Cheiranthus cheiri, 98
Chimerae, 177, 261, 267
Chromatin rearrangements, 312
Chromosomes, as large chain mole-
cules, 300, 310

loss of whole, 147

natural races, 152

sticky, 280
Chrysomelid beetles, 251
Cirsium, 278

Coccinellids, color pattern of, 247
Colias edusa, 130, 250
Coloboma, 40

Conditioned dominance, 107
Cotton, dominigenes, 108
Crack patterns, 261
Crataegomespilus, 177
Crataegus, 178
Crepis, 149

369

370

I'HYSlOUXiKAL (1KNETIC8

Critical period (.sec Sensitive period:

Crossovers, 292

Crustacea, asymmetry of claws, 227

Cubitus interrupt us, 312

Cvcurbita ptpo, 217

Cyclopia, 47

Cynthia moth, experiments on wing

pattern, 247
Cytoplasm, diseased, 269

influence of, 56

and nucleus, 263

and sex, 278
Cytoplasmic genes, 279
Cytoplasmic heredity in mosses, 271
Cytoplasmic influence upon gene-
controlled characters, 270

1)

Dahlia, flower color, 169

Dahlia variabilis, 98

Datura, 151, 213

Dauermodification, 276-277

Deer horns, 77

Deficiencies, 133, 147, 176, 287, 298,

307, 313
Deletions, 147

Determination stream, 193, 195-196,
198, 200, 230, 239-240, 245,
253, 255, 261
Development, potentialities of, 3
Diffusion, of gene-controlled prod-
ucts, 180

of growth substances, 190
Dogs, spotting in, 258
Dominance, 65, 99

change of, 99, 112, 161

coefficient, 104-105

and development, 99

and environment, 102

phylogenetic theory, 109, 121-122

temperature shocks, 107

theory of, 100

and threshold, 103
Dominigenes, 62, 107, 129, 306, 312
Dosage, 312

changes, 137, 151

compensation, 130, 131

Dosage, differences, 154, 285
and whole chromosomes, 147
by X-chromosomes, 125
Drosophila funebris, asymmetry,

223
dominigenes, 109
maternal inheritance, 207
phenocopies, 5
venae transversae, 71
Drosophila melanogaster, asymme-
try, 222, 225, 230
Bar-eye series, 11, 31-32, 44, 46,
73, 76, 89, 104, 106, 120, 128,
135, 137, 139, 147, 154, 167,
215, 284, 297, 304-305, 312
bobbed series, 118, 136, 145
bristle pattern, 240
bristles, 19, 46, 222
deficiency, 299

See also Deficiencies
dominance, 116
dominigenes, 22, 107-108
duplication, 137, 147, 307, 313
eggs, 280
eye-colors, 32, 34, 91, 115, 132,

164, 173, 175, 183, 307
eye-mutants, 26, 30, 34, 75
fourth chromosome, 147, 149, 151
haploid tissue, 44
hybrids, 108
intersexes, 125, 189
larval characters, 207
legs, double, 70
maternal inheritance, 207
mosaics, 177, 182
mutants, abdomen rotatum, 225

abnormal abdomen, 70, 134

achaete, 199

aristapedia, 36, 208

beaded, 192

bithorax, 208

chubby, 38

claret, 175

crippled, 70

cubitus interruptus, 306

Dichaete, 73, 85

dumpy, 231

expanded, 191, 212

SUBJECT INDEX

371

Drosophila melanogaster, mutants,
Facet-notch, 297
giant, 36
Gull, 109
Lobe, 114, 164
lozenge, 86

miniature, 176, 191, 212
Minute, 46, 162, 164, 175, 313
Paletranslocation, 132
Plum, 307-308
proboscipedia, 208
short wing, 81
tetraptera, 70, 208
truncate, 87, 116, 191, 230
origin of mutants, 313
pattern, 228
phenocopies, 5, 9, 45, 56
pigment races, 91
pleiotropy, 81
rates, 73

Russian race, 123
Scute series, 72, 116, 235, 237, 239,

295, 303, 306, 312
sensitive periods, 76
sex-differences, 132
species, 66

temperature effect (see pheno-
copies, Bar, scute, vestigial)
transplantation, 117, 133, 185, 187
unstable genes, 290
vestigial series, 12, 17, 25, 26, 30,
42-43, 46, 54, 56, 58, 62, 66,
69, 74, 76, 79, 83-84, 87, 103-
104, 109, 118, 123, 129, 139,
190, 216, 222, 226, 228, 238
wing, 161, 212, 217, 234
Drosophila obscura, 86, 107, 165
Drosophila pseudoobscura, 207
Drosophila simulans, 108, 183
Drosophila virilis, 176, 290
Drosophila willistoni, 86

E

Embryology, experimental, 46
Embryonic fields, 179, 202, 204
Embryonic segregation, 204

Ephestia kuhniella, 10, 20-21, 23, 45,
79, 127, 183-184, 196, 233, 243-
244, 246-247, 250-251, 253, 261

Epilachnia chrysomelina, 247

Epilobium, 274-275, 278-279

Episomes, 297

Epistatic minimum, 66, 68

Eumelanin, 34

Evocator, 202-205, 221

Exaggeration, 133, 137

Expressivity, 66

Eye, human, anomalies of, 41

Feather, development, 34, 218

growth, 218

pigment, 218, 261
Ferns, variegation in, 291
Fetalization, 41
Fiddler crabs, claws of, 221
Fishes, pattern of pigment, 259
Flax, male sterility in, 279
Flour moth (see Ephestia kuhniella)
Flower colors, chemistry of, 94

dominance, 114

interaction, 169
Fowl, creeper, 25, 214

dominigenes, 109

frizzle, 107

rumpless, 10, 25-26, 46, 107, 214
Frog, 179
Funaria, 153
Fundulus, 47

G

Galeopsis tetrahit, 55
Gammarus, 34, 39, 46, 53, 206
Gastropacha, 233
Gene, action in cell, 264

activation, 263, 265

aggregate, 295

autocatalyst, 282

balance, 133, 152

chemical models, 299

chemical processes controlling, 99

372

PHYSIOLOGICAL GENETICS

Gene, com pared in haploids, diploids,
tetraploids, 148

cytoplasmic control, 279-2S0

developmental control, 205

different doses, 123

heterozygous, 99

interaction, 1">C>

loss of chromosome material, 298

molecule, 283

nature of, 282

pattern, 209

part of higher unit, 299

physics, 293

pleiotropy, 287

potency, 285

products of hormone type, 182

protoplasmic, 275

quantity, 136, 301

sick, 289-291, 295

side-chain theory, 300

size, 283

theory of tomorrow, 303

unstable, 288, 295
Genie environment, 107
( ienomere hypothesis, 295
Germinal vesicle, bursting of, 264
Gluthatione and growth, 193
Goldfish, 260
Gonad and eye color, 183
Gradients, theory of, 203
Growth, by cell division, 212
by differentiation, 213
by increase in cell size, 212, 216
Growth hormone, 188
Guinea pigs, 49, 66, 70, 73, 82, 87,

118, 226, 258
Gynandromorphism, 10, 124, 130,

174-175, 181, 183, 261
Gypsy moth (see Lymantria dispar)

II

Habrobracon, 26, 164, 177, 183, 233,

273
Harmozones, 182
Helix, 112
Hemmungsmissbildung, 39

Beterogonic growth, 167, 210-211,

215, 217 2 IS, 221
Homoeosis, 36, 116, 208, 313
Hormones, 181-182

of pupation, 38
Horse, spotting in, 258
Hypomorphic gene, 103, 134

Identical twins, 222, 226
Idiochromatin, 264

Induction, 177, 179-181, 201

Inert regions of chromosome, 284

Insect development, 181

Interaction of the genes, 156-157,
162

Internal genetic environment, 156

Intersexuality, 39, 40, 52, 66, 68,
124, 148, 172, 181, 188-189,
194-195, 229, 231, 238, 241, 285

Intrachromosomal balance, 299

Inversions, 306-307

Invisibles, 312

K

Karyoplasmic ratio, 152

L

Larval characters, 207

Lathyrus, flower color of, 114

Lebistes, 261

Lepidoptera (see Butterflies)

Leptinotarsa, 5, 99

Lethality, 42-44, 181, 298

Liesegang rings, 204-205, 220, 252-

253, 256-257, 261
Linum usitalissimum, 40
Lipochromes, 36

Lymantria dispar, caterpillars, 52-
53, 72, 82, 114, 207, 251

cytoplasmic heredity, 271, 273,
279

dominance, 112-113

geographic races, 55, 126, 270

growth, 77

SUBJECT INDEX

373

Lymantria dispar, hypcrmales, 69

induction, 190

intersexuality, 39, 40, 52, 63, 67,
181, 189, 194, 222-223, 226,
238

molting, 77, 127

mutants, 247, 250

pattern, 228-229, 231-232, 241

phenocopy, 40, 233

prothetely, 173

sex in development, 131

sex-controlled inheritance, 129

sex-genes, 63, 72, 118, 124

sex- races, 145

size inheritance, 213

threshold, 117, 181
Lymantria monacha, 128, 207, 250
Lymnaeiis, asymmetry, 224, 227

M

Mutant, identical with a develop-
mental stage, 40
Mutant gene, type of reaction con-
trolled by, 174
Mutants, phenocopies and the de-
velopment of, 23-51
Mutation, excitation, 294
rate, 284, 289, 312
reverse, 289-290
theory, 285, 288, 292
by X-rays, 175, 293, 312

X

Nematodes, 212
Neomorphic, 137

Nucleus and cytoplasm, interaction
of, 263

0

Maize, 62, 94, 99, 115, 148-149, 280,

303
Malva parviflora, 291
Mammals, abnormalities, 24-25, 43

agouti hair, 220

chemistry of color, 91

piebald, 258
Mass mutation, 313
Maternal inheritance, 184, 207, 277,

280
Matthiola, 280
Melanism, 210, 233, 250
Melopsittacus undulatus, 35
Merogony, 180, 267-268
Mespilus, 177-178
Modifiers, 168, 312
Molluscs, right- and left-hand coil-
ing, 206, 280
Molting hormone, 77, 221
Monsters, production of, 46
Mosaics, 174, 239
Mosses, species crosses in, 278
Mouse, 24, 25, 44, 49-51, 108, 115,

161, 188, 214-215, 258
Multiple alleles, 12, 81, 84, 91, 114,
154, 191, 290, 293, 299, 303, 313
Multiple factors, 150, 161, 286

Oats, color of, 150
Oenothera, 63-64, 69, 269
Optical colors of wing scales, 248
Organ-forming stuffs, 179, 205
Organizer, 190, 201-202, 208
Oryzias, 23, 259-260
Otocephaly, 50
Outlets of pigment, 245, 253, 258

Papaver Rhoeas, 98

Papilio centralis, 4

Papilio dardanus, 130, 249, 255

Papilio machaon, 4

Papilio mimetic, 251

Papilio podalirius, 4

pupal wing of, 233
Papilio polymorphic species, 234
Papilio polytes, 130
Papilio sphyrus, 4
Papilio, tails on hind wings, 234
Papilio zanclaeus, 4
Parasemia, 250
Pattern, composite, 241

crack, 261

374

PHYSIOLOGICAL GENETICS

Pattern, development, 178, 200-201,
205, 208, 210, 243, 261
gradient, 48
growth, 210, 262

localisation. 245, '-'IS

unit ant genes, 25 I
pigment, 193, L96, 258-259

zebra, 2<>1

Teas, rogues in, 291
Pelargonium sonale, 98, 169
Penetrance, 66, 69, 71

Petunia, critical period for, 22

Pharbitw Nil, maple alleles, 86

Phenocopy, 4, 9-11, 14, 16, 18, 22-
23, 26, 40, 44-46, 49, 58, 66, 83,
102, 106, 128, 196, 215, 235, 238,
243, 253

Phenogenetics, 23

Pheomelanin, 34-35

Physcomitrella, 153

Phylogeny, 311

Pholiota mutabilis, 268

Pigmentation, chemical models, 92

Pigment cells in newts, 179

Plasmon, 56, 275-276

Plastids, 269, 281

Platypoecilus, 44, 260

Pleiotropy, 78, 115

Plymouth Rock feathers, 219, 261

Pollen fertility, 279

Pollen sterility, 274, 280

Polydactyly in guinea pigs, asym-
metry, 226

Polymorphism, nielanic, 250

Polyploidy, 152, 153, 169, 213

Position effect, 109, 138, 292, 297,
300, 302, 304, 306, 308-309, 312

Potato beetle, color varieties of, 91

Potencies of genes, 285

Poultry, extra toes in, 226

Presence-absence, 123

Primula sinensis, 80, 98, 187

Promethea moth, 228

Proteinase, 315

Prothetely, 173

Protosome, 297

Psammechinus, 267

Psychid moths, L53

Pterylae, development of, 259, 261

B

Rabbit, 54, 86, 91, 155, 160-161,

258
Kai, 19, 225, 258

Rate, of growth and hormones in
leathers, 219

of segmentation, 208
Hate concept, 51
Rate genes, 52, 155
Reactions within the cells, 174
Rearrangements, spontaneous, 312
Rodents, 108, 157, 160

colors of the hair of, 157

S

Salivary chromosomes, 284, 307

Sumiu cecropia, 243

Satureia, 278

Saturnid moth, 246

Scardafetta inca, 5

Scliizophyllum commune, 268

Sea-urchin egg, 203

Secale, 38

Selachia, ovocytes of, 264

Self-differentiation, 201

Semilethals, 298

Sensitive period, 8, 13-22, 30, 44,

73, 75, 87, 128, 198, 240, 246-

247
Sepalody, 65

Serological specificity, 99, 275
Sex chromosomes, dosage through,

130
Sex-controlled phenotype, 125, 129
Sex determination, 66, 124
Sex-determining reactions, 172
Sex differentiation in vertebrates,

181
Sex genes, cytoplasmic influence

upon, 279
intracellular action of, 178
Sex hormones, 181, 183, 188, 190,

219

SUBJECT INDEX

375

Sexual polymorphism, 249
Shape, hereditary control of, 217
Siamese cats, 258
Silkworm, 53, 77, 115, US, 127,

206-207, 251, 280
Size, different, transplantation of

organs between animals of, 220
hereditary, 193, 213, 217
Snails, dextral and sinistral, 277
Somatic crossing over, 176
Somatic mutations, 176, 225, 201
Somatic segregation, mosaic spots

by, 176
Species hybrids of Lepidoptera, 189
Spilosoma lubricipeda, 250
Step allelomorphism, 295, 299
Stereoisomers, 310
Stratification, 205, 220, 240, 252,

256, 262, 266
Symmetry, 221

Teleost crosses, 49

Temperature action, 4-5, 12, 22, 46,

75, 139, 233, 312
Tenebrio, 113, 173
Tetraploidy, 147, 150
Tetrasomics, 147
Thais polyxenn, 242
Thermophenes, 104, 143
Threshold effect, 12, 65, 71, 89, 115,

117, 120, 123, 127-128, 135, 150,

163, 180, 188, 190, 219, 223, 297
Tiger stripes, 257
Tobacco, 112, 278
Tomato, 212, 218
Translocation, 137, 145, 307, 313
Transplantation of organs between

animals of different hereditary

size, 220
Triploidy, 124, 147, 149
Trisomies, 147, 213

Triticum, 38

Triton (Triturus), 179-180, 268

Trophochromatin, 264

Tumor, hereditary, 271
Tumors, melanotic, in fish hybrids,
44

U

Unequal crossing over, 304
Unstable genes, 175

Van't Hoff formula, 44
Vanessa ichnusa, 4
Vanessa levana-prorsa, 249
Vanessa polaris, 4
Vanessa urticae, 4, 6-7
Variegation, 289

in ferns, 291
Variegation, Nicotiana, 292
Vertebral ankylosis, 25
Vertebral column, hereditary abnor-
malities of, 78
Vertebrates, monstrosities in, 81
Vicia faba, 272

W

Wasp patterns, 257
Wave mechanics, 288
Wheat, development of ear form in,
38

Y-chromosome, 279
Z

Zebra stripes, 257

c:;^c< ***&£